LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic...

41
Aydi, E, Orio, M, Beardmore, AP, Ness, J-U, Page, KL, Kuin, NPM, Walter, FM, Buckley, DAH, Mohamed, S, Whitelock, P, Osborne, JP, Strader, J, Chomiuk, L, Darnley, MJ, Dobrotka, A, Kniazev, A, Miszalski, B, Myers, G, Ospina, N, Henze, M, Starrfield, S and Woodward, CE Multiwavelength observations of V407 Lupi (ASASSN-16kt) --- a very fast nova erupting in an intermediate polar http://researchonline.ljmu.ac.uk/8965/ Article LJMU has developed LJMU Research Online for users to access the research output of the University more effectively. Copyright © and Moral Rights for the papers on this site are retained by the individual authors and/or other copyright owners. Users may download and/or print one copy of any article(s) in LJMU Research Online to facilitate their private study or for non-commercial research. You may not engage in further distribution of the material or use it for any profit-making activities or any commercial gain. The version presented here may differ from the published version or from the version of the record. Please see the repository URL above for details on accessing the published version and note that access may require a subscription. For more information please contact [email protected] http://researchonline.ljmu.ac.uk/ Citation (please note it is advisable to refer to the publisher’s version if you intend to cite from this work) Aydi, E, Orio, M, Beardmore, AP, Ness, J-U, Page, KL, Kuin, NPM, Walter, FM, Buckley, DAH, Mohamed, S, Whitelock, P, Osborne, JP, Strader, J, Chomiuk, L, Darnley, MJ, Dobrotka, A, Kniazev, A, Miszalski, B, Myers, G, Ospina, N, Henze, M, Starrfield, S and Woodward, CE (2018) LJMU Research Online

Transcript of LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic...

Page 1: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

Aydi E Orio M Beardmore AP Ness J-U Page KL Kuin NPM Walter FM

Buckley DAH Mohamed S Whitelock P Osborne JP Strader J Chomiuk L

Darnley MJ Dobrotka A Kniazev A Miszalski B Myers G Ospina N Henze

M Starrfield S and Woodward CE

Multiwavelength observations of V407 Lupi (ASASSN-16kt) --- a very fast nova

erupting in an intermediate polar

httpresearchonlineljmuacuk8965

Article

LJMU has developed LJMU Research Online for users to access the research output of the

University more effectively Copyright copy and Moral Rights for the papers on this site are retained by

the individual authors andor other copyright owners Users may download andor print one copy of

any article(s) in LJMU Research Online to facilitate their private study or for non-commercial research

You may not engage in further distribution of the material or use it for any profit-making activities or

any commercial gain

The version presented here may differ from the published version or from the version of the record

Please see the repository URL above for details on accessing the published version and note that

access may require a subscription

For more information please contact researchonlineljmuacuk

httpresearchonlineljmuacuk

Citation (please note it is advisable to refer to the publisherrsquos version if you

intend to cite from this work)

Aydi E Orio M Beardmore AP Ness J-U Page KL Kuin NPM Walter

FM Buckley DAH Mohamed S Whitelock P Osborne JP Strader J

Chomiuk L Darnley MJ Dobrotka A Kniazev A Miszalski B Myers G

Ospina N Henze M Starrfield S and Woodward CE (2018)

LJMU Research Online

httpresearchonlineljmuacuk

Mon Not R Astron Soc 000 1ndash28 (2018) Printed 3 July 2018 (MN LATEX style file v22)

Multiwavelength observations of V407 Lupi(ASASSN-16kt) mdash a very fast nova erupting in anintermediate polar

E Aydi12⋆ M Orio34 A P Beardmore5 J-U Ness6 K L Page5 N P M Kuin7

F M Walter8 D A H Buckley1 S Mohamed12 P Whitelock12 J P Osborne5

J Strader9 L Chomiuk9 M J Darnley10 A Dobrotka11 A Kniazev11213

B Miszalski112 G Myers14 N Ospina15 M Henze16 S Starrfield17

and C E Woodward18

1South African Astronomical Observatory PO Box 9 7935 Observatory South Africa2Department of Astronomy University of Cape Town Private Bag X3 Rondebosch 7701 South Africa3INAFndashOsservatorio di Padova vicolo dell Osservatorio 5 I-35122 Padova Italy4Department of Astronomy University of Wisconsin 475 N Charter Str Madison WI 53704 USA5X-ray and Observational Astronomy Group Department of Physics amp Astronomy University of Leicester LE1 7RH UK6XMM-Newton Observatory SOC European Space Astronomy Centre Camino Bajo del Castillo sn Urb Villafranca del CastilloE-28692 Villanueva de la Canada Madrid Spain7University College London Mullard Space Science Laboratory Holmbury St Mary Dorking RH5 6NT UK8Department of Physics and Astronomy Stony Brook University Stony Brook NY 11794-3800 USA9Center for Data Intensive and Time Domain Astronomy Department of Physics and Astronomy Michigan State University East LansingMI 4882410Astrophysics Research Institute Liverpool John Moores University IC2 Liverpool Science Park Liverpool L3 5RF UK11Advanced Technologies Research Institute Slovak University of Technology in Bratislava Paulinska 16 91724 Trnava Slovak Republic12Southern African Large Telescope Foundation PO Box 9 Observatory 7935 South Africa13Special Astrophysical Observatory of RAS Nizhnij Arkhyz Karachai-Circassia 369167 Russia14American Association of Variable Star Observers 5 Inverness Way Hillsborough CA 9401015Via Pietro Pomponazzi 33 35124 Padova Italy16Department of Astronomy San Diego State University San Diego CA 92182 USA17School of Earth and Space Exploration Arizona State University Tempe Arizona 85287-1404 USA18Minnesota Institute for Astrophysics University of Minnesota Minneapolis MN 55455 USA

Accepted 2018 July 1 Received 2018 June 8 in original form 2018 April 17

ABSTRACT

We present a detailed study of the 2016 eruption of nova V407 Lupi (ASASSN-16kt)including optical near-infrared X-ray and ultraviolet data from SALT SMARTSSOAR Chandra Swift and XMM-Newton Timing analysis of the multiwavelengthlight-curves shows that from 168 days post-eruption and for the duration of the X-ray supersoft source phase two periods at 565 s and 357 h are detected We suggestthat these are the rotational period of the white dwarf and the orbital period ofthe binary respectively and that the system is likely to be an intermediate polarThe optical light-curve decline was very fast (t2 6 29 d) suggesting that the whitedwarf is likely massive (amp 125M⊙) The optical spectra obtained during the X-raysupersoft source phase exhibit narrow complex and moving emission lines of He iialso characteristics of magnetic cataclysmic variables The optical and X-ray datashow evidence for accretion resumption while the X-ray supersoft source is still onpossibly extending its duration

Key words stars individual (V407 Lup) ndash novae cataclysmic variables ndash whitedwarfs

arX

iv1

807

0070

6v1

[as

tro-

phS

R]

2 J

ul 2

018

2 Aydi et al

1 INTRODUCTION

Classical novae (CNe) are stellar eruptions that take placewithin the surface layers of accreting white dwarfs (WDs)in cataclysmic variable (CV) systems These are interactingbinaries consisting of a WD primary accreting from a sec-ondary that typically fills its Roche lobe (see eg Warner1995) The hydrogen-rich material accreted by the WD ac-cumulates on its surface causing an increase in pressure anddensity to a level where a thermonuclear runaway (TNR) istriggered (Starrfield 1989 Starrfield Iliadis amp Hix 2008 Joseamp Shore 2008 Starrfield Iliadis amp Hix 2016) Due to thisa part of the accreted envelope is ejected The fast expand-ing envelope moves together with the optical photosphere(Hachisu amp Kato 2006 2014) in what is known as the ldquofire-ball stagerdquo During this the brightness of the star increasesby 8 up to 15 mag (Payne-Gaposchkin 1964) and reaches itsmaximum visual brightness when the optical photosphereattains its maximum radius (Warner 1995 Hachisu amp Kato2014) Following the TNR the remaining hydrogen-rich ac-creted envelope continuously burns on the WD surface andmay become visible when the expanding ejecta become op-tically thin

After maximum the optical depth of the expandingejecta decreases progressively and therefore the optical pho-tosphere starts shrinking This shifts the emission to higherenergies and would peak in the soft X-ray band and a Su-perSoft Source (SSS) emerges unless the ejecta become op-tically thin before that happens in which case the SSS emer-gence is due to the visibility of the WD photosphere whichis at that time extended due to the ongoing nuclear burning(Krautter 2008 Osborne 2015) As the hydrogen in the sur-face layers is consumed the SSS ldquoturns offrdquo and the systemeventually evolves back to quiescence (for a series of reviewsof CNe see Bode amp Evans 2008)

Observing CNe across many wavelengths is essential toprovide a full picture of the eruption and its characteris-tics such as the different physical parameters (eg tem-peratures and velocities) the gas ejection mechanisms theradiative processes the nuclear reactions the shocks in theejectawind the dust formation and the properties of theprogenitor However despite their importance systematicmultiwavelength studies are still lacking and only a few CNehave been followed panchromatically and in detail (see egShore et al 2013 Schwarz et al 2015 Bode et al 2016 Shoreet al 2016 Darnley et al 2016 Li et al 2017 Mason et al2018 Aydi et al 2018) Therefore observing these eventsin multiple frequency bands is essential for a better under-standing of these individual events and ultimately adding tothe wealth of knowledge in the field

While most nova eruptions occur in non-magnetic CVswhere the WD is usually weakly magnetized (magnetic fieldof the WD is 106 G) a few novae have been seen to occurin magnetic CVs (mCVs) These systems form a sub-groupof CVs in which the WD is highly magnetized and they in-clude two main sub-types polars which have the strongestmagnetic fields (amp 107 G) and intermediate polars (IPs) withweaker fields (106 ndash 107 G) The strong magnetic field of po-lars causes the spin (rotational) period of the WD and theorbital period of the binary to synchronize This is not thecase for IPs where the optical X-ray and ultraviolet (UV)light-curves might show multiple periodicities modulated

on the orbital period of the binary (with typical values of3 to 6 h) the spin period of the WD (with values rangingbetween sim 05 up to sim 70min) and the sideband periods(typically dominated by the beat period) Although in suchsystems the mass-accretion mechanism is dominated at somepoint by the magnetic field of the WD the accretion ontothe surface of WD can still result in a nova eruption Fora comprehensive review of mCVs and their mass-transfer-accretion mechanisms see Warner (1995) and Hellier (2001)

The only nova known to have erupted in a polar sys-tem is V1500 Cyg (Kaluzny amp Semeniuk 1987 StockmanSchmidt amp Lamb 1988) The eruption which occurred in1975 was one of the most luminous on record along with CPPup (Warner 1985) and nova SMCN 2016-10a (Aydi et al2018) On the other hand a few novae have been either con-firmed (GK Per DQ Her V4743 Sgr and Nova Scorpii 1437AD) or suggested (V533 Her DD Cir V1425 Aql M31N2007-12b and V2491 Cyg) to occur in IPs (see eg Os-borne et al 2001 King Osborne amp Schenker 2002 WarnerampWoudt 2002 Bianchini et al 2003 Woudt ampWarner 20032004 Leibowitz et al 2006 Dobrotka amp Ness 2010 Pietschet al 2011 Zemko et al 2016 Potter amp Buckley 2018 andreferences therein) The effect of the magnetic field on thenova eruption and progress is not well understood

In this paper we present a multiwavelength studyof nova V407 Lupi which was discovered by the All-SkyAutomated Survey for Supernovae (ASAS-SN)1 on HJD24576555 (2016 September 240 UT Stanek et al 2016)at V = 91 and is located at equatorial coordinates of(α δ)J20000 = (15h29m0182s ndash4449prime40primeprime89) and Galacticcoordinates of (l b) = (33009 9573) The last ASAS-SNpre-discovery observation reported the source at V gt 155on HJD 24576515 Therefore in the following we assumeHJD 24576555 (2016 September 240 UT) is t0 (eruptionstart) Our study consists of a set of comprehensive multi-wavelength observations obtained with the aim of studyingin detail the post-eruption behaviour of the nova at opti-cal near-infrared (NIR) ultraviolet (UV) and X-ray wave-lengths Note that this object has not been seen in eruptionbefore This paper is structured as follows in Section 2 wepresent the observations and data reduction The analysisof the photometric and spectroscopic results are given inSections 3 and 4 respectively We present the discussion inSection 5 while Section 6 contains a summary and the con-clusions

2 OBSERVATIONS AND DATA REDUCTION

21 SMARTS observations and data reduction

Since 2016 September 26 (day 2) the eruption has beenmonitored using the Small and Moderate Aperture ResearchTelescope System (SMARTS) to obtain optical BVRI andNIR JHK photometric observations The integration timesat JHK were 15 s (three 5 s dithered images) before 2016October 6 (day 12) and 30 s thereafter Optical observationsare single images of 30 s 25 s 20 s and 20 s integrations re-spectively in BVRI prior to 2016 October 6 (day 12) and

1 httpwwwastronomyohio-stateeduasassnindexshtml

nova V407 Lup 3

uniformly 50 s thereafter The procedure followed for reduc-ing the SMARTS photometry is detailed in Walter et al(2012) Fig B1 represents the SMARTS BVRI and JHK

observations See Table D1 for a log of the observations

22 SALT high-resolution Echelle spectroscopy

We used the SALT High Resolution Spectrograph (HRSBarnes et al 2008 Bramall et al 2010 Bramall et al 2012Crause et al 2014) to obtain observations on the nights of2017 March 6 April 20 May 31 June 13 29 July 05 24 29and August 06 (respectively days 164 209 250 263 279285 304 309 and 317) HRS a dual beam fibre-fed Echellespectrograph housed in a temperature stabilized vacuumtank was used in the low-resolution (LR) mode to obtainall the observations This mode provides two spectral rangesblue (3800 ndash 5550 A) and red (5450 ndash 9000 A) at a resolutionof R sim 15000 A weekly set of HRS calibrations includingfour ThAr + Ar arc spectra and four spectral flats is ob-tained in all the modes (low medium and high-resolution)All the HRS observations are 1800 s exposures See Table D2for a log of the observations

The primary reduction was conducted using the SALTscience pipeline (Crawford et al 2010) which includes over-scan correction bias subtraction and gain correction Therest of the reduction and relative flux calibration wasdone using the standard MIDAS FEROS (Stahl Kaufer ampTubbesing 1999) and echelle (Ballester 1992) packages Thereduction procedure is described in detail in Kniazev Gvara-madze amp Berdnikov (2016) Note that absolute flux calibra-tion is not feasible with SALT data As part of the SALTdesign the effective area of the telescope constantly changeswith the moving pupil during the track and exposures

23 SOAR medium-resolution spectroscopy

We performed optical spectroscopy of the nova on 2017 June20 (day 270) using the Goodman spectrograph (ClemensCrain amp Anderson 2004) on the 41m Southern Astrophys-ical Research (SOAR) telescope We obtained a single ex-posure of 600 s using a 600 lmmminus1 grating and a 107primeprime slitto provide a resolution of sim 13 A over the range of 3200 ndash6800 A The spectrum was reduced using the standard rou-tine and a relative flux calibration has been applied

We also obtained two spectra using the Goodman spec-trograph on the nights of 2018 January 20 and 21 (days484 and 485) each of 20min exposure For these two ob-servations we used a setup with a 2100 lmmminus1 grating anda 095primeprimeslit yielding a resolution of about 088 A full widthat half maximum (FWHM) (56 km sminus1) over a wavelengthrange of 4280 ndash 4950 AA The spectra were reduced and op-timally extracted in the usual manner

24 Swift X-ray and UV observations

241 Swift XRT observations

The Neil Gehrels Swift Observatory (hereafter Swift Gehrels et al 2004) first observed V407 Lup on 2016 Septem-ber 26 two days after the discovery The high optical bright-ness of the nova at this time meant that the UVOpticalTelescope (UVOT Roming et al 2005) could not be used

and the X-ray Telescope (XRT Burrows et al 2005) neededto be operated in Windowed Timing (WT) mode for theinitial observation in order to minimize the effects of op-tical loading whereby a large number of optical photonscan pile-up to appear as a false X-ray signature2 No X-raysource was detected at this time (lt 002 count sminus1 grade 0ndash single pixel events) or in the individual subsequent dailyobservations taken in Photon Counting (PC) mode between2016 October 2 and 9 (days 8 ndash 15) Coadding these PC datathere is a source detected at the 99 per cent confidence levelwith a count rate of (14+05

minus04) times10minus3 count sminus1 (grade 0)There was also a suggestion of a detection in the October10 (day 16) data alone These detections could not be con-firmed at a higher significance before V407 Lup became tooclose to the Sun for Swift to observe on 2016 October 13(day 19)

Observations recommenced on 2017 February 21 (day150) finding a bright supersoft X-ray source (Beardmoreet al 2017) with a count rate of 561plusmn 03 count sminus1 Bythis time the UVOT could also be safely operated and a UVsource with uvw2 = 1349plusmn 002mag was measured Dailyobservations were performed between 2017 February 22 and28 (days 151 ndash 157) followed by observations approximatelyevery two to three days until 2017 April 24 (day 212) Obser-vations were continued though with a slightly lower cadenceof every four days until 2017 September 10 (day 351) withthe cadence then decreasing to every eight days until 2017October 13 (day 384) when the Sun constraint began againA final dataset was obtained once the source again emergedfrom the Sun constraint on 2018 January 23 (day 486) Allobservations were typically 05-1 ks in duration with the ex-ception of the final observation which was 3 ks In addition tothis regular monitoring a high cadence campaign was per-formed between 2017 July 4 ndash 8 (days 283 ndash 287) aimed atpinning down the periodicity seen in the UVOT light-curve(see Section 34) In this case sim 500 s snapshots were takenapproximately every 6 h Because of complications causedby a nearby bright star no UVOT data were actually col-lected and so the campaign was repeated from 2017 July28 ndashAugust 1 (days 307 ndash 311) using an offset pointing toavoid this problem The Swift XRT observations log is givenin Table D3

The Swift data were processed with the standardHEASoft tools (version 620)3 and analysed using themost up-to-date calibration files All observations be-tween 2017 February 21 and August 11 (days 150 and321 post-eruption) inclusive were obtained using WTmode because of the high brightness of the X-raysource These data were extracted using a circular re-gion of a radius of 20 pixels (1 pixel = 236 arcsec)for the source and a background annulus as describedat httpwwwswiftacukanalysisxrtbackscalphp Ob-servations from 2017 August 15 (day 325) onwards weretaken with the XRT in PC mode the first three of thesesuffered from pile-up so an annulus (outer radius 30 pix-els inner exclusion radius decreasing from seven to three

2 httpwwwswiftacukanalysisxrtoptical_loading

php3 httpsheasarcgsfcnasagovFTPsoftwareftools

releasearchiveRelease_Notes_620

4 Aydi et al

pixels) was used when extracting the source counts Thelater data were analysed using circular regions decreasingin radius from 20 to 10 pixels as the source further fadedBackground counts were estimated from near-by source-freecircular regions of 60 pixels radius

The X-ray spectra were binned to provide a minimum of1 count binminus1 to facilitate Cash statistic (Cash 1979) fittingwithin xspec (Arnaud 1996)

242 Swift UVOT observations

Observations with the Swift UVOT instrument startedonce the nova could be observed again after its passage be-hind the Sun On 2017 February 26 (day 156) the nova wasobserved in the uvw2 (λcentral = 1928 A) filter the next dayin the uvm2 (λcentral = 2246 A) filter Photometric observa-tions continued until 2017 October 13 (day 385) A log ofthe observations is given in Table D3

Weekly observations with the UV grism started on 2017March 7 until April 23 (day 165 until day 212) when theywere discontinued as the brightness was too low The UVgrism provides a spectral range of sim 1700 ndash 5100 A Thelog of the observations are given in Table D4 The eightgrism observations were reduced using the UVOTPY soft-ware (Kuin 2014) using the calibration described in Kuinet al (2015) with a recent update to the sensitivity affect-ing the response of the UV grism mainly below 2000 A 4The net continuum counts in the UV part of the individualspectra are very low so the first four and last four spectrawere summed which also improved the signal to noise (SN)of the weaker lines significantly No reddening correction hasbeen applied to the spectra

25 XMM-Newton X-ray and UV observations

V407 Lup was observed by XMM-Newton from 2017 March11 1145 to 1708 UT 1685 days post-eruption with anexposure duration of 23 ks (Ness et al 2017) The XMM-

Newton observatory consists of five different X-ray in-struments behind three mirrors plus an optical monitor(OMMason et al 2001 Talavera 2009) which all observesimultaneously For a full description of the X-ray instru-ments onboard of XMM-Newton see Jansen et al (2001)den Herder et al (2001) Struder et al (2001) Turner et al(2001) and Aschenbach (2002)

In this work we only used the spectra and light-curvesfrom the Reflection Grating Spectrometer (RGSden Herderet al 2001) the light-curves from the European PhotonImaging Camera (EPIC)pn (Struder et al 2001) and theOM light-curvesgrism spectra (see Table D5 for a log of theobservations)

The OM took five exposures one of which with the vis-ible grism provided a spectrum between 3000 and 6000 A(XMM-Newton could not take a U -grism spectrum becauseof a contaminating nearby star so we only obtained a V -grism spectrum) and four with Science User Defined imag-ing plus fast mode which provided a UV light-curve withthe uvw1 (λ sim 2910plusmn 830 A) filter

4 seehttpmssluclacuknpmkGrism

The RGS provides a spectral range of 6 ndash 38 A fully cov-ering the Wien tail of SSS spectra (30 ndash 80 eV) while theRayleigh Jeans tail is not visible (owing to interstellar ab-sorption) It also provides X-ray light-curves in the sameenergy range RGS consists of two instruments RGS1 andRGS2 One of the nine CCDs in RGS2 is suffering from sometechnical problems We Combine the RGS1 and RGS2 spec-tra in order to overcome the problem of the broken CCD inthe RGS2 instrument The EPICpn instrument was oper-ated in Timing Mode with medium filter providing anotherX-ray light-curve

26 Chandra X-ray observations

V407 Lup was observed with the Chandra X-ray Observa-

tory (hereafter Chandra Weisskopf et al 2000) using theHigh Resolution Camera (HRC Murray et al 1997) andthe Low Energy Transmission Grating (LETG Brinkmanet al 2000ab) on 2018 August 30 (day 3406) with an ex-posure time of 34 ks The LETG instrument provides a spec-trum over a range of sim 12 ndash 175 A (007 ndash 1033 keV) how-ever most of the flux from nova V407 Lup is in the range15 ndash 50 A (024 ndash 082 keV) The light-curve provided by theHRC instrument has a range of 006 ndash 10 keV

3 PHOTOMETRIC RESULTS AND ANALYSIS

31 Optical light-curve parameters

Several parameters characterize nova light-curves includingthe rise rate the rise time to maximum light the maximumlight the decline rate and the decline behaviour (see egHounsell et al 2010 Cao et al 2012) Nova V407 Lup wasnot extensively observed during its rise to maximum

Based on V -band and visual (Vis) measurements fromSMARTS and the American Association of Variable StarObservers (AAVSO)5 we see that the nova reached V =68 04 d after t0 Then it was reported at V = 64 on HJD245765624 (Prieto 2016) and Vis = 63 on HJD 245765657reaching Vis = 56 on HJD 245765690 (see Table C1) Halfa day later the nova was at V = 633 and Vis = 65 thendropped to V = 785 in around two days indicating thatthe decline had started Hence we assume HJD 245765690as tmax (day 14)

Caution is required in interpreting magnitudes of no-vae This is particularly so for Vis where Hα may con-tribute to the flux detected by eye but not to CCD V -magnitudes Furthermore normal transformation relationsfor CCD magnitudes cannot be used for objects with strongemission lines Unfortunately no spectra were taken aroundmaximum light so the contribution and development of lineemission is unknown until later times (day 5 Izzo et al2018) In the following we assume that maximum light wasVmax 6 60 on HJD 245765690 Since novae do not usu-ally have strong line emission at maximum light (van denBergh amp Younger 1987) and given that simultaneous V andVis measurements show very similar magnitudes before andafter maximum light it may be that V sim Vis =56 onHJD 245765690 is a reasonable estimate However we use

5 httpswwwaavsoorg

nova V407 Lup 5

Figure 1 V -band light-curve using data from SMARTS (green circles) AAVSO (red squares) and ATels from Stanek et al (2016)(yellow circle) Chen et al (2016) (blue stars) and Prieto (2016) (cyan diamond) A zoom-in plot around the first sim 25 days is added forclarity Within this plot the blue dashed line represents a power-law fit to the data (before day 25) excluding the discovery measurementthe black horizontal dashed lines represent from top to bottom Vmax Vmax + 2 and Vmax + 3 respectively the black vertical dashedlines represent from left to right tmax t2 and t3 respectively

Vmax 6 60 for the rest of the analysis to avoid overestimat-ing the maximum brightness

We construct a V -band light-curve by combining thepublished photometry from different telescopes and instru-ments and the SMARTS data (see Fig 1) noting the abovecaveat This shows a rapid rise and a sharp peak followed bya fast decline Such behaviour is characteristic of the S-classnova light-curves (Strope Schaefer amp Henden 2010) whichis considered stereotypical for a nova However the broad-band light-curves show day-to day variability after t3 similarto that of an O-class light-curves (see Fig B1) It is worthnoting that there is no presence for a plateau in the tail ofthe light-curve (after sim 3 ndash 6mag from maximum) similar tothat seen in recurrent novae (see Schaefer 2010)

Although there is a gap in the light-curve between day26 and day 107 (due to solar constraints) we fit a power-lawto the light-curve (only early decline - before day 25) andderive t2 6 29plusmn 05 d t3 6 68 plusmn 10 d and a power index ofsim 0166 We took into account the use of measurements fromdifferent instruments by increasing (times 2) the uncertainty ont2 and t3

With t2 6 29 d and a decline rate of sim 069mag dminus1

over t2 nova V407 Lup is a ldquovery-fastrdquo nova in the classifi-

6 Izzo et al (2018) have derived longer values of t2 and t3 This

was due to misinterpretation of their light-curve (private commu-nication with L Izzo)

Figure 2 The UVOT uvw2 (λcentral = 1928 A) light-curve plot-ted against day since t0

cation of Payne-Gaposchkin (1964) and is one of the fastestknown examples Only a few other novae have shown adecline time t2 30 days including M31N 2008-12a USco V1500 Cyg V838 Her V394 CrA and V4160 Sgr (seeegYoung et al 1976 Schaefer 2010 Munari et al 2011and table 5 in Darnley et al 2016)

6 Aydi et al

Figure 3 First panel the XMM-Newton EPICpn X-ray light-curve Second panel the RGS1 X-ray light-curve The X-ray light-curvesshow two broad (sim 2 ndash 3 ks wide) dips separated by 126 ks consistent with the 357 h modulation seen in the Swift-UVOT data (seeSection 34) Third panel the uvw1 OM fast mode (black) and imaging mode (red) light-curves The first OM exposure is a grismspectrum hence the light-curves only start at sim 5 ks The blue line represents a best fit sine curve to the OM data with a period of376plusmn03 h This period is consistent within uncertainty with the period seen in the Swift-UVOT data (see Section 34 for further

discussion on the timing analysis) The y-axis in all the panels are counts per second

32 The Swift UVOT light-curve

The UVOT uvw2 light-curve (Fig 2) declined slowly afterday 150 before flattening off at around a uvw2 magnitude of142 by around day 200 with signs of a 01mag variabilitysuperimposed on the overall decline We find periodicities inthe light-curve which we discuss in Section 34 The UV datarebrightened after day 300 to reach a uvw2 magnitude of137 around day 320 Then the brightness started to declineagain reaching a uvw2 magnitude of 149 at day sim 385Fig B2 shows a direct comparison between the Swift X-raySwift UV and SMARTS optical light-curves

33 The XMM-Newton and Chandra UV and

X-ray light-curves

The XMM-Newton RGS1 X-ray light-curve the EPICpnX-ray light-curve and the OM UV fast and imaging modeslight-curves are all presented in Fig 3 The XMM-Newton

RGS1 light-curve shows evidence of variability on differenttimescales This includes two broad (sim 2 ndash 3 ks wide) dipsseparated by sim 126 ks sim 35 h The OM fast mode UV light-curve shows a variation characterized by a 376plusmn03 h periodwhich is also consistent (within uncertainty) with the dura-tion between the dips in the RGS1 light-curve Howeverno other periodicity is found in the OM light-curve TheChandra HRC light-curve (Fig 4) also shows evidence forshort-term variability which we discuss in Section 34

Figure 4 The Chandra High Resolution Camera (HRC) X-ray

light-curve The HRC energy range is 006 ndash 10 keV There is aclear short-term variability in the light-curve and it is discussed

in Section 34

34 Timing analysis

Since the SwiftUVOT uvw2 light-curve (Fig 2) shows evi-dence for intrinsic variability superimposed on the long termdecline we searched for periodicities in the data Fig 5shows a Lomb-Scargle periodogram (LSP) of the barycen-tric corrected uvw2 light-curve after subtracting a third or-der polynomial to remove the long term trend The mostsignificant peaks occur at periods of 35731 plusmn 00014 h and

nova V407 Lup 7

Figure 5 Upper panel the Lomb Scargle periodogram of thedetrended uvw2 light-curve revealing significant modulation atfrequencies of 67168 dminus1 (777times10minus5 sminus1 P = 3573 h) and217696 dminus1 (252 times 10minus4 sminus1 P = 1102 h with estimated un-certainties of 00003 dminus1) marked with red dashed lines Lowerpanel periodogram of the observing window showing Swift rsquos or-

bital frequency (dotted line)

11024plusmn 00002 h which are aliases of each other caused bySwift rsquos 16 h orbit One of these periods might represent theorbital period of the binary (Porb) The low cadence of theUVOT data means we cannot break the degeneracy betweenthe periods (the high-cadence observation interval betweendays 307 and 311 failed to resolve this issue) The amplitudeof the modulation is approximately 01mag

We performed a similar Lomb-Scargle analysis of theXMM-Newton RGS1 and Chandra HRC data (see Fig 6)The light-curve of the former shows considerable flux vari-ations while the flux in the latter is approximately con-stant over the duration of the observation The XMM-

Newton periodogram is dominated by a low frequency ofΩ = 777times10minus5 sminus1 (357 h) This signal only covers sim twocycles of the 357 h period and cannot be believed in iso-lation However as this signal is consistent with the longperiod seen in the UVOT data we identify it as the same357 h modulation

The long-timescale modulation and variability seen inthe XMM-Newton light-curves the time between the dipsand the weak modulation observed in the OM light-curve(see Figs 3 and 6) are all consistent with the 357 h signalseen in the Swift UV light-curve In addition the absenceof any variability in the XMM-Newton RGS and OM light-curves related to the 11 h signal (seen in the Swift UV light-curve) has the potential to rule out the modulation at thisshort period Therefore we assume that the 357 h is mostlikely the Porb of the binary

Ignoring this long-term variability seen in the low fre-quency part of XMM-Newton periodogram we find signifi-cant signals at ω ≃ 177times10minus3 sminus1 in both (XMM-Newton

and Chandra) light-curves More precisely the periodicitieswe found are at 56504plusmn 068 s for the Chandra HRC light-curve and 56496plusmn 050 s for the XMM-Newton RGS light-curve The SwiftXRT observations are almost all too shortto usefully constrain the modulation at this short period

Figure 6 Lomb Scargle periodograms of the XMM-NewtonRGS1 (top) and the Chandra HRC (bottom) light-curves Vari-ous frequencies are indicated where Ω = 1Porb and ω = 1Pspin

(see text for more details) We present zoom-in plots around the177times10minus3 sminus1 in each periodogram

Figure 7 Top from top to bottom soft RGS1 light-curve hard

RGS1 light-curve and the hardness ratio The energy bandsfor the soft and hard light-curves are 235 ndash 370 A (03325 ndash

0528 keV) and 150 ndash 235 A (0528 ndash 0827 keV) respectively Bot-tom same as top but all folded at the 565 s spin period andnormalized to their respective mean rates Therefore the y-axisrepresents the relative amplitude of the modulation in both thesoft and hard bands

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 2: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

httpresearchonlineljmuacuk

Mon Not R Astron Soc 000 1ndash28 (2018) Printed 3 July 2018 (MN LATEX style file v22)

Multiwavelength observations of V407 Lupi(ASASSN-16kt) mdash a very fast nova erupting in anintermediate polar

E Aydi12⋆ M Orio34 A P Beardmore5 J-U Ness6 K L Page5 N P M Kuin7

F M Walter8 D A H Buckley1 S Mohamed12 P Whitelock12 J P Osborne5

J Strader9 L Chomiuk9 M J Darnley10 A Dobrotka11 A Kniazev11213

B Miszalski112 G Myers14 N Ospina15 M Henze16 S Starrfield17

and C E Woodward18

1South African Astronomical Observatory PO Box 9 7935 Observatory South Africa2Department of Astronomy University of Cape Town Private Bag X3 Rondebosch 7701 South Africa3INAFndashOsservatorio di Padova vicolo dell Osservatorio 5 I-35122 Padova Italy4Department of Astronomy University of Wisconsin 475 N Charter Str Madison WI 53704 USA5X-ray and Observational Astronomy Group Department of Physics amp Astronomy University of Leicester LE1 7RH UK6XMM-Newton Observatory SOC European Space Astronomy Centre Camino Bajo del Castillo sn Urb Villafranca del CastilloE-28692 Villanueva de la Canada Madrid Spain7University College London Mullard Space Science Laboratory Holmbury St Mary Dorking RH5 6NT UK8Department of Physics and Astronomy Stony Brook University Stony Brook NY 11794-3800 USA9Center for Data Intensive and Time Domain Astronomy Department of Physics and Astronomy Michigan State University East LansingMI 4882410Astrophysics Research Institute Liverpool John Moores University IC2 Liverpool Science Park Liverpool L3 5RF UK11Advanced Technologies Research Institute Slovak University of Technology in Bratislava Paulinska 16 91724 Trnava Slovak Republic12Southern African Large Telescope Foundation PO Box 9 Observatory 7935 South Africa13Special Astrophysical Observatory of RAS Nizhnij Arkhyz Karachai-Circassia 369167 Russia14American Association of Variable Star Observers 5 Inverness Way Hillsborough CA 9401015Via Pietro Pomponazzi 33 35124 Padova Italy16Department of Astronomy San Diego State University San Diego CA 92182 USA17School of Earth and Space Exploration Arizona State University Tempe Arizona 85287-1404 USA18Minnesota Institute for Astrophysics University of Minnesota Minneapolis MN 55455 USA

Accepted 2018 July 1 Received 2018 June 8 in original form 2018 April 17

ABSTRACT

We present a detailed study of the 2016 eruption of nova V407 Lupi (ASASSN-16kt)including optical near-infrared X-ray and ultraviolet data from SALT SMARTSSOAR Chandra Swift and XMM-Newton Timing analysis of the multiwavelengthlight-curves shows that from 168 days post-eruption and for the duration of the X-ray supersoft source phase two periods at 565 s and 357 h are detected We suggestthat these are the rotational period of the white dwarf and the orbital period ofthe binary respectively and that the system is likely to be an intermediate polarThe optical light-curve decline was very fast (t2 6 29 d) suggesting that the whitedwarf is likely massive (amp 125M⊙) The optical spectra obtained during the X-raysupersoft source phase exhibit narrow complex and moving emission lines of He iialso characteristics of magnetic cataclysmic variables The optical and X-ray datashow evidence for accretion resumption while the X-ray supersoft source is still onpossibly extending its duration

Key words stars individual (V407 Lup) ndash novae cataclysmic variables ndash whitedwarfs

arX

iv1

807

0070

6v1

[as

tro-

phS

R]

2 J

ul 2

018

2 Aydi et al

1 INTRODUCTION

Classical novae (CNe) are stellar eruptions that take placewithin the surface layers of accreting white dwarfs (WDs)in cataclysmic variable (CV) systems These are interactingbinaries consisting of a WD primary accreting from a sec-ondary that typically fills its Roche lobe (see eg Warner1995) The hydrogen-rich material accreted by the WD ac-cumulates on its surface causing an increase in pressure anddensity to a level where a thermonuclear runaway (TNR) istriggered (Starrfield 1989 Starrfield Iliadis amp Hix 2008 Joseamp Shore 2008 Starrfield Iliadis amp Hix 2016) Due to thisa part of the accreted envelope is ejected The fast expand-ing envelope moves together with the optical photosphere(Hachisu amp Kato 2006 2014) in what is known as the ldquofire-ball stagerdquo During this the brightness of the star increasesby 8 up to 15 mag (Payne-Gaposchkin 1964) and reaches itsmaximum visual brightness when the optical photosphereattains its maximum radius (Warner 1995 Hachisu amp Kato2014) Following the TNR the remaining hydrogen-rich ac-creted envelope continuously burns on the WD surface andmay become visible when the expanding ejecta become op-tically thin

After maximum the optical depth of the expandingejecta decreases progressively and therefore the optical pho-tosphere starts shrinking This shifts the emission to higherenergies and would peak in the soft X-ray band and a Su-perSoft Source (SSS) emerges unless the ejecta become op-tically thin before that happens in which case the SSS emer-gence is due to the visibility of the WD photosphere whichis at that time extended due to the ongoing nuclear burning(Krautter 2008 Osborne 2015) As the hydrogen in the sur-face layers is consumed the SSS ldquoturns offrdquo and the systemeventually evolves back to quiescence (for a series of reviewsof CNe see Bode amp Evans 2008)

Observing CNe across many wavelengths is essential toprovide a full picture of the eruption and its characteris-tics such as the different physical parameters (eg tem-peratures and velocities) the gas ejection mechanisms theradiative processes the nuclear reactions the shocks in theejectawind the dust formation and the properties of theprogenitor However despite their importance systematicmultiwavelength studies are still lacking and only a few CNehave been followed panchromatically and in detail (see egShore et al 2013 Schwarz et al 2015 Bode et al 2016 Shoreet al 2016 Darnley et al 2016 Li et al 2017 Mason et al2018 Aydi et al 2018) Therefore observing these eventsin multiple frequency bands is essential for a better under-standing of these individual events and ultimately adding tothe wealth of knowledge in the field

While most nova eruptions occur in non-magnetic CVswhere the WD is usually weakly magnetized (magnetic fieldof the WD is 106 G) a few novae have been seen to occurin magnetic CVs (mCVs) These systems form a sub-groupof CVs in which the WD is highly magnetized and they in-clude two main sub-types polars which have the strongestmagnetic fields (amp 107 G) and intermediate polars (IPs) withweaker fields (106 ndash 107 G) The strong magnetic field of po-lars causes the spin (rotational) period of the WD and theorbital period of the binary to synchronize This is not thecase for IPs where the optical X-ray and ultraviolet (UV)light-curves might show multiple periodicities modulated

on the orbital period of the binary (with typical values of3 to 6 h) the spin period of the WD (with values rangingbetween sim 05 up to sim 70min) and the sideband periods(typically dominated by the beat period) Although in suchsystems the mass-accretion mechanism is dominated at somepoint by the magnetic field of the WD the accretion ontothe surface of WD can still result in a nova eruption Fora comprehensive review of mCVs and their mass-transfer-accretion mechanisms see Warner (1995) and Hellier (2001)

The only nova known to have erupted in a polar sys-tem is V1500 Cyg (Kaluzny amp Semeniuk 1987 StockmanSchmidt amp Lamb 1988) The eruption which occurred in1975 was one of the most luminous on record along with CPPup (Warner 1985) and nova SMCN 2016-10a (Aydi et al2018) On the other hand a few novae have been either con-firmed (GK Per DQ Her V4743 Sgr and Nova Scorpii 1437AD) or suggested (V533 Her DD Cir V1425 Aql M31N2007-12b and V2491 Cyg) to occur in IPs (see eg Os-borne et al 2001 King Osborne amp Schenker 2002 WarnerampWoudt 2002 Bianchini et al 2003 Woudt ampWarner 20032004 Leibowitz et al 2006 Dobrotka amp Ness 2010 Pietschet al 2011 Zemko et al 2016 Potter amp Buckley 2018 andreferences therein) The effect of the magnetic field on thenova eruption and progress is not well understood

In this paper we present a multiwavelength studyof nova V407 Lupi which was discovered by the All-SkyAutomated Survey for Supernovae (ASAS-SN)1 on HJD24576555 (2016 September 240 UT Stanek et al 2016)at V = 91 and is located at equatorial coordinates of(α δ)J20000 = (15h29m0182s ndash4449prime40primeprime89) and Galacticcoordinates of (l b) = (33009 9573) The last ASAS-SNpre-discovery observation reported the source at V gt 155on HJD 24576515 Therefore in the following we assumeHJD 24576555 (2016 September 240 UT) is t0 (eruptionstart) Our study consists of a set of comprehensive multi-wavelength observations obtained with the aim of studyingin detail the post-eruption behaviour of the nova at opti-cal near-infrared (NIR) ultraviolet (UV) and X-ray wave-lengths Note that this object has not been seen in eruptionbefore This paper is structured as follows in Section 2 wepresent the observations and data reduction The analysisof the photometric and spectroscopic results are given inSections 3 and 4 respectively We present the discussion inSection 5 while Section 6 contains a summary and the con-clusions

2 OBSERVATIONS AND DATA REDUCTION

21 SMARTS observations and data reduction

Since 2016 September 26 (day 2) the eruption has beenmonitored using the Small and Moderate Aperture ResearchTelescope System (SMARTS) to obtain optical BVRI andNIR JHK photometric observations The integration timesat JHK were 15 s (three 5 s dithered images) before 2016October 6 (day 12) and 30 s thereafter Optical observationsare single images of 30 s 25 s 20 s and 20 s integrations re-spectively in BVRI prior to 2016 October 6 (day 12) and

1 httpwwwastronomyohio-stateeduasassnindexshtml

nova V407 Lup 3

uniformly 50 s thereafter The procedure followed for reduc-ing the SMARTS photometry is detailed in Walter et al(2012) Fig B1 represents the SMARTS BVRI and JHK

observations See Table D1 for a log of the observations

22 SALT high-resolution Echelle spectroscopy

We used the SALT High Resolution Spectrograph (HRSBarnes et al 2008 Bramall et al 2010 Bramall et al 2012Crause et al 2014) to obtain observations on the nights of2017 March 6 April 20 May 31 June 13 29 July 05 24 29and August 06 (respectively days 164 209 250 263 279285 304 309 and 317) HRS a dual beam fibre-fed Echellespectrograph housed in a temperature stabilized vacuumtank was used in the low-resolution (LR) mode to obtainall the observations This mode provides two spectral rangesblue (3800 ndash 5550 A) and red (5450 ndash 9000 A) at a resolutionof R sim 15000 A weekly set of HRS calibrations includingfour ThAr + Ar arc spectra and four spectral flats is ob-tained in all the modes (low medium and high-resolution)All the HRS observations are 1800 s exposures See Table D2for a log of the observations

The primary reduction was conducted using the SALTscience pipeline (Crawford et al 2010) which includes over-scan correction bias subtraction and gain correction Therest of the reduction and relative flux calibration wasdone using the standard MIDAS FEROS (Stahl Kaufer ampTubbesing 1999) and echelle (Ballester 1992) packages Thereduction procedure is described in detail in Kniazev Gvara-madze amp Berdnikov (2016) Note that absolute flux calibra-tion is not feasible with SALT data As part of the SALTdesign the effective area of the telescope constantly changeswith the moving pupil during the track and exposures

23 SOAR medium-resolution spectroscopy

We performed optical spectroscopy of the nova on 2017 June20 (day 270) using the Goodman spectrograph (ClemensCrain amp Anderson 2004) on the 41m Southern Astrophys-ical Research (SOAR) telescope We obtained a single ex-posure of 600 s using a 600 lmmminus1 grating and a 107primeprime slitto provide a resolution of sim 13 A over the range of 3200 ndash6800 A The spectrum was reduced using the standard rou-tine and a relative flux calibration has been applied

We also obtained two spectra using the Goodman spec-trograph on the nights of 2018 January 20 and 21 (days484 and 485) each of 20min exposure For these two ob-servations we used a setup with a 2100 lmmminus1 grating anda 095primeprimeslit yielding a resolution of about 088 A full widthat half maximum (FWHM) (56 km sminus1) over a wavelengthrange of 4280 ndash 4950 AA The spectra were reduced and op-timally extracted in the usual manner

24 Swift X-ray and UV observations

241 Swift XRT observations

The Neil Gehrels Swift Observatory (hereafter Swift Gehrels et al 2004) first observed V407 Lup on 2016 Septem-ber 26 two days after the discovery The high optical bright-ness of the nova at this time meant that the UVOpticalTelescope (UVOT Roming et al 2005) could not be used

and the X-ray Telescope (XRT Burrows et al 2005) neededto be operated in Windowed Timing (WT) mode for theinitial observation in order to minimize the effects of op-tical loading whereby a large number of optical photonscan pile-up to appear as a false X-ray signature2 No X-raysource was detected at this time (lt 002 count sminus1 grade 0ndash single pixel events) or in the individual subsequent dailyobservations taken in Photon Counting (PC) mode between2016 October 2 and 9 (days 8 ndash 15) Coadding these PC datathere is a source detected at the 99 per cent confidence levelwith a count rate of (14+05

minus04) times10minus3 count sminus1 (grade 0)There was also a suggestion of a detection in the October10 (day 16) data alone These detections could not be con-firmed at a higher significance before V407 Lup became tooclose to the Sun for Swift to observe on 2016 October 13(day 19)

Observations recommenced on 2017 February 21 (day150) finding a bright supersoft X-ray source (Beardmoreet al 2017) with a count rate of 561plusmn 03 count sminus1 Bythis time the UVOT could also be safely operated and a UVsource with uvw2 = 1349plusmn 002mag was measured Dailyobservations were performed between 2017 February 22 and28 (days 151 ndash 157) followed by observations approximatelyevery two to three days until 2017 April 24 (day 212) Obser-vations were continued though with a slightly lower cadenceof every four days until 2017 September 10 (day 351) withthe cadence then decreasing to every eight days until 2017October 13 (day 384) when the Sun constraint began againA final dataset was obtained once the source again emergedfrom the Sun constraint on 2018 January 23 (day 486) Allobservations were typically 05-1 ks in duration with the ex-ception of the final observation which was 3 ks In addition tothis regular monitoring a high cadence campaign was per-formed between 2017 July 4 ndash 8 (days 283 ndash 287) aimed atpinning down the periodicity seen in the UVOT light-curve(see Section 34) In this case sim 500 s snapshots were takenapproximately every 6 h Because of complications causedby a nearby bright star no UVOT data were actually col-lected and so the campaign was repeated from 2017 July28 ndashAugust 1 (days 307 ndash 311) using an offset pointing toavoid this problem The Swift XRT observations log is givenin Table D3

The Swift data were processed with the standardHEASoft tools (version 620)3 and analysed using themost up-to-date calibration files All observations be-tween 2017 February 21 and August 11 (days 150 and321 post-eruption) inclusive were obtained using WTmode because of the high brightness of the X-raysource These data were extracted using a circular re-gion of a radius of 20 pixels (1 pixel = 236 arcsec)for the source and a background annulus as describedat httpwwwswiftacukanalysisxrtbackscalphp Ob-servations from 2017 August 15 (day 325) onwards weretaken with the XRT in PC mode the first three of thesesuffered from pile-up so an annulus (outer radius 30 pix-els inner exclusion radius decreasing from seven to three

2 httpwwwswiftacukanalysisxrtoptical_loading

php3 httpsheasarcgsfcnasagovFTPsoftwareftools

releasearchiveRelease_Notes_620

4 Aydi et al

pixels) was used when extracting the source counts Thelater data were analysed using circular regions decreasingin radius from 20 to 10 pixels as the source further fadedBackground counts were estimated from near-by source-freecircular regions of 60 pixels radius

The X-ray spectra were binned to provide a minimum of1 count binminus1 to facilitate Cash statistic (Cash 1979) fittingwithin xspec (Arnaud 1996)

242 Swift UVOT observations

Observations with the Swift UVOT instrument startedonce the nova could be observed again after its passage be-hind the Sun On 2017 February 26 (day 156) the nova wasobserved in the uvw2 (λcentral = 1928 A) filter the next dayin the uvm2 (λcentral = 2246 A) filter Photometric observa-tions continued until 2017 October 13 (day 385) A log ofthe observations is given in Table D3

Weekly observations with the UV grism started on 2017March 7 until April 23 (day 165 until day 212) when theywere discontinued as the brightness was too low The UVgrism provides a spectral range of sim 1700 ndash 5100 A Thelog of the observations are given in Table D4 The eightgrism observations were reduced using the UVOTPY soft-ware (Kuin 2014) using the calibration described in Kuinet al (2015) with a recent update to the sensitivity affect-ing the response of the UV grism mainly below 2000 A 4The net continuum counts in the UV part of the individualspectra are very low so the first four and last four spectrawere summed which also improved the signal to noise (SN)of the weaker lines significantly No reddening correction hasbeen applied to the spectra

25 XMM-Newton X-ray and UV observations

V407 Lup was observed by XMM-Newton from 2017 March11 1145 to 1708 UT 1685 days post-eruption with anexposure duration of 23 ks (Ness et al 2017) The XMM-

Newton observatory consists of five different X-ray in-struments behind three mirrors plus an optical monitor(OMMason et al 2001 Talavera 2009) which all observesimultaneously For a full description of the X-ray instru-ments onboard of XMM-Newton see Jansen et al (2001)den Herder et al (2001) Struder et al (2001) Turner et al(2001) and Aschenbach (2002)

In this work we only used the spectra and light-curvesfrom the Reflection Grating Spectrometer (RGSden Herderet al 2001) the light-curves from the European PhotonImaging Camera (EPIC)pn (Struder et al 2001) and theOM light-curvesgrism spectra (see Table D5 for a log of theobservations)

The OM took five exposures one of which with the vis-ible grism provided a spectrum between 3000 and 6000 A(XMM-Newton could not take a U -grism spectrum becauseof a contaminating nearby star so we only obtained a V -grism spectrum) and four with Science User Defined imag-ing plus fast mode which provided a UV light-curve withthe uvw1 (λ sim 2910plusmn 830 A) filter

4 seehttpmssluclacuknpmkGrism

The RGS provides a spectral range of 6 ndash 38 A fully cov-ering the Wien tail of SSS spectra (30 ndash 80 eV) while theRayleigh Jeans tail is not visible (owing to interstellar ab-sorption) It also provides X-ray light-curves in the sameenergy range RGS consists of two instruments RGS1 andRGS2 One of the nine CCDs in RGS2 is suffering from sometechnical problems We Combine the RGS1 and RGS2 spec-tra in order to overcome the problem of the broken CCD inthe RGS2 instrument The EPICpn instrument was oper-ated in Timing Mode with medium filter providing anotherX-ray light-curve

26 Chandra X-ray observations

V407 Lup was observed with the Chandra X-ray Observa-

tory (hereafter Chandra Weisskopf et al 2000) using theHigh Resolution Camera (HRC Murray et al 1997) andthe Low Energy Transmission Grating (LETG Brinkmanet al 2000ab) on 2018 August 30 (day 3406) with an ex-posure time of 34 ks The LETG instrument provides a spec-trum over a range of sim 12 ndash 175 A (007 ndash 1033 keV) how-ever most of the flux from nova V407 Lup is in the range15 ndash 50 A (024 ndash 082 keV) The light-curve provided by theHRC instrument has a range of 006 ndash 10 keV

3 PHOTOMETRIC RESULTS AND ANALYSIS

31 Optical light-curve parameters

Several parameters characterize nova light-curves includingthe rise rate the rise time to maximum light the maximumlight the decline rate and the decline behaviour (see egHounsell et al 2010 Cao et al 2012) Nova V407 Lup wasnot extensively observed during its rise to maximum

Based on V -band and visual (Vis) measurements fromSMARTS and the American Association of Variable StarObservers (AAVSO)5 we see that the nova reached V =68 04 d after t0 Then it was reported at V = 64 on HJD245765624 (Prieto 2016) and Vis = 63 on HJD 245765657reaching Vis = 56 on HJD 245765690 (see Table C1) Halfa day later the nova was at V = 633 and Vis = 65 thendropped to V = 785 in around two days indicating thatthe decline had started Hence we assume HJD 245765690as tmax (day 14)

Caution is required in interpreting magnitudes of no-vae This is particularly so for Vis where Hα may con-tribute to the flux detected by eye but not to CCD V -magnitudes Furthermore normal transformation relationsfor CCD magnitudes cannot be used for objects with strongemission lines Unfortunately no spectra were taken aroundmaximum light so the contribution and development of lineemission is unknown until later times (day 5 Izzo et al2018) In the following we assume that maximum light wasVmax 6 60 on HJD 245765690 Since novae do not usu-ally have strong line emission at maximum light (van denBergh amp Younger 1987) and given that simultaneous V andVis measurements show very similar magnitudes before andafter maximum light it may be that V sim Vis =56 onHJD 245765690 is a reasonable estimate However we use

5 httpswwwaavsoorg

nova V407 Lup 5

Figure 1 V -band light-curve using data from SMARTS (green circles) AAVSO (red squares) and ATels from Stanek et al (2016)(yellow circle) Chen et al (2016) (blue stars) and Prieto (2016) (cyan diamond) A zoom-in plot around the first sim 25 days is added forclarity Within this plot the blue dashed line represents a power-law fit to the data (before day 25) excluding the discovery measurementthe black horizontal dashed lines represent from top to bottom Vmax Vmax + 2 and Vmax + 3 respectively the black vertical dashedlines represent from left to right tmax t2 and t3 respectively

Vmax 6 60 for the rest of the analysis to avoid overestimat-ing the maximum brightness

We construct a V -band light-curve by combining thepublished photometry from different telescopes and instru-ments and the SMARTS data (see Fig 1) noting the abovecaveat This shows a rapid rise and a sharp peak followed bya fast decline Such behaviour is characteristic of the S-classnova light-curves (Strope Schaefer amp Henden 2010) whichis considered stereotypical for a nova However the broad-band light-curves show day-to day variability after t3 similarto that of an O-class light-curves (see Fig B1) It is worthnoting that there is no presence for a plateau in the tail ofthe light-curve (after sim 3 ndash 6mag from maximum) similar tothat seen in recurrent novae (see Schaefer 2010)

Although there is a gap in the light-curve between day26 and day 107 (due to solar constraints) we fit a power-lawto the light-curve (only early decline - before day 25) andderive t2 6 29plusmn 05 d t3 6 68 plusmn 10 d and a power index ofsim 0166 We took into account the use of measurements fromdifferent instruments by increasing (times 2) the uncertainty ont2 and t3

With t2 6 29 d and a decline rate of sim 069mag dminus1

over t2 nova V407 Lup is a ldquovery-fastrdquo nova in the classifi-

6 Izzo et al (2018) have derived longer values of t2 and t3 This

was due to misinterpretation of their light-curve (private commu-nication with L Izzo)

Figure 2 The UVOT uvw2 (λcentral = 1928 A) light-curve plot-ted against day since t0

cation of Payne-Gaposchkin (1964) and is one of the fastestknown examples Only a few other novae have shown adecline time t2 30 days including M31N 2008-12a USco V1500 Cyg V838 Her V394 CrA and V4160 Sgr (seeegYoung et al 1976 Schaefer 2010 Munari et al 2011and table 5 in Darnley et al 2016)

6 Aydi et al

Figure 3 First panel the XMM-Newton EPICpn X-ray light-curve Second panel the RGS1 X-ray light-curve The X-ray light-curvesshow two broad (sim 2 ndash 3 ks wide) dips separated by 126 ks consistent with the 357 h modulation seen in the Swift-UVOT data (seeSection 34) Third panel the uvw1 OM fast mode (black) and imaging mode (red) light-curves The first OM exposure is a grismspectrum hence the light-curves only start at sim 5 ks The blue line represents a best fit sine curve to the OM data with a period of376plusmn03 h This period is consistent within uncertainty with the period seen in the Swift-UVOT data (see Section 34 for further

discussion on the timing analysis) The y-axis in all the panels are counts per second

32 The Swift UVOT light-curve

The UVOT uvw2 light-curve (Fig 2) declined slowly afterday 150 before flattening off at around a uvw2 magnitude of142 by around day 200 with signs of a 01mag variabilitysuperimposed on the overall decline We find periodicities inthe light-curve which we discuss in Section 34 The UV datarebrightened after day 300 to reach a uvw2 magnitude of137 around day 320 Then the brightness started to declineagain reaching a uvw2 magnitude of 149 at day sim 385Fig B2 shows a direct comparison between the Swift X-raySwift UV and SMARTS optical light-curves

33 The XMM-Newton and Chandra UV and

X-ray light-curves

The XMM-Newton RGS1 X-ray light-curve the EPICpnX-ray light-curve and the OM UV fast and imaging modeslight-curves are all presented in Fig 3 The XMM-Newton

RGS1 light-curve shows evidence of variability on differenttimescales This includes two broad (sim 2 ndash 3 ks wide) dipsseparated by sim 126 ks sim 35 h The OM fast mode UV light-curve shows a variation characterized by a 376plusmn03 h periodwhich is also consistent (within uncertainty) with the dura-tion between the dips in the RGS1 light-curve Howeverno other periodicity is found in the OM light-curve TheChandra HRC light-curve (Fig 4) also shows evidence forshort-term variability which we discuss in Section 34

Figure 4 The Chandra High Resolution Camera (HRC) X-ray

light-curve The HRC energy range is 006 ndash 10 keV There is aclear short-term variability in the light-curve and it is discussed

in Section 34

34 Timing analysis

Since the SwiftUVOT uvw2 light-curve (Fig 2) shows evi-dence for intrinsic variability superimposed on the long termdecline we searched for periodicities in the data Fig 5shows a Lomb-Scargle periodogram (LSP) of the barycen-tric corrected uvw2 light-curve after subtracting a third or-der polynomial to remove the long term trend The mostsignificant peaks occur at periods of 35731 plusmn 00014 h and

nova V407 Lup 7

Figure 5 Upper panel the Lomb Scargle periodogram of thedetrended uvw2 light-curve revealing significant modulation atfrequencies of 67168 dminus1 (777times10minus5 sminus1 P = 3573 h) and217696 dminus1 (252 times 10minus4 sminus1 P = 1102 h with estimated un-certainties of 00003 dminus1) marked with red dashed lines Lowerpanel periodogram of the observing window showing Swift rsquos or-

bital frequency (dotted line)

11024plusmn 00002 h which are aliases of each other caused bySwift rsquos 16 h orbit One of these periods might represent theorbital period of the binary (Porb) The low cadence of theUVOT data means we cannot break the degeneracy betweenthe periods (the high-cadence observation interval betweendays 307 and 311 failed to resolve this issue) The amplitudeof the modulation is approximately 01mag

We performed a similar Lomb-Scargle analysis of theXMM-Newton RGS1 and Chandra HRC data (see Fig 6)The light-curve of the former shows considerable flux vari-ations while the flux in the latter is approximately con-stant over the duration of the observation The XMM-

Newton periodogram is dominated by a low frequency ofΩ = 777times10minus5 sminus1 (357 h) This signal only covers sim twocycles of the 357 h period and cannot be believed in iso-lation However as this signal is consistent with the longperiod seen in the UVOT data we identify it as the same357 h modulation

The long-timescale modulation and variability seen inthe XMM-Newton light-curves the time between the dipsand the weak modulation observed in the OM light-curve(see Figs 3 and 6) are all consistent with the 357 h signalseen in the Swift UV light-curve In addition the absenceof any variability in the XMM-Newton RGS and OM light-curves related to the 11 h signal (seen in the Swift UV light-curve) has the potential to rule out the modulation at thisshort period Therefore we assume that the 357 h is mostlikely the Porb of the binary

Ignoring this long-term variability seen in the low fre-quency part of XMM-Newton periodogram we find signifi-cant signals at ω ≃ 177times10minus3 sminus1 in both (XMM-Newton

and Chandra) light-curves More precisely the periodicitieswe found are at 56504plusmn 068 s for the Chandra HRC light-curve and 56496plusmn 050 s for the XMM-Newton RGS light-curve The SwiftXRT observations are almost all too shortto usefully constrain the modulation at this short period

Figure 6 Lomb Scargle periodograms of the XMM-NewtonRGS1 (top) and the Chandra HRC (bottom) light-curves Vari-ous frequencies are indicated where Ω = 1Porb and ω = 1Pspin

(see text for more details) We present zoom-in plots around the177times10minus3 sminus1 in each periodogram

Figure 7 Top from top to bottom soft RGS1 light-curve hard

RGS1 light-curve and the hardness ratio The energy bandsfor the soft and hard light-curves are 235 ndash 370 A (03325 ndash

0528 keV) and 150 ndash 235 A (0528 ndash 0827 keV) respectively Bot-tom same as top but all folded at the 565 s spin period andnormalized to their respective mean rates Therefore the y-axisrepresents the relative amplitude of the modulation in both thesoft and hard bands

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 3: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

Mon Not R Astron Soc 000 1ndash28 (2018) Printed 3 July 2018 (MN LATEX style file v22)

Multiwavelength observations of V407 Lupi(ASASSN-16kt) mdash a very fast nova erupting in anintermediate polar

E Aydi12⋆ M Orio34 A P Beardmore5 J-U Ness6 K L Page5 N P M Kuin7

F M Walter8 D A H Buckley1 S Mohamed12 P Whitelock12 J P Osborne5

J Strader9 L Chomiuk9 M J Darnley10 A Dobrotka11 A Kniazev11213

B Miszalski112 G Myers14 N Ospina15 M Henze16 S Starrfield17

and C E Woodward18

1South African Astronomical Observatory PO Box 9 7935 Observatory South Africa2Department of Astronomy University of Cape Town Private Bag X3 Rondebosch 7701 South Africa3INAFndashOsservatorio di Padova vicolo dell Osservatorio 5 I-35122 Padova Italy4Department of Astronomy University of Wisconsin 475 N Charter Str Madison WI 53704 USA5X-ray and Observational Astronomy Group Department of Physics amp Astronomy University of Leicester LE1 7RH UK6XMM-Newton Observatory SOC European Space Astronomy Centre Camino Bajo del Castillo sn Urb Villafranca del CastilloE-28692 Villanueva de la Canada Madrid Spain7University College London Mullard Space Science Laboratory Holmbury St Mary Dorking RH5 6NT UK8Department of Physics and Astronomy Stony Brook University Stony Brook NY 11794-3800 USA9Center for Data Intensive and Time Domain Astronomy Department of Physics and Astronomy Michigan State University East LansingMI 4882410Astrophysics Research Institute Liverpool John Moores University IC2 Liverpool Science Park Liverpool L3 5RF UK11Advanced Technologies Research Institute Slovak University of Technology in Bratislava Paulinska 16 91724 Trnava Slovak Republic12Southern African Large Telescope Foundation PO Box 9 Observatory 7935 South Africa13Special Astrophysical Observatory of RAS Nizhnij Arkhyz Karachai-Circassia 369167 Russia14American Association of Variable Star Observers 5 Inverness Way Hillsborough CA 9401015Via Pietro Pomponazzi 33 35124 Padova Italy16Department of Astronomy San Diego State University San Diego CA 92182 USA17School of Earth and Space Exploration Arizona State University Tempe Arizona 85287-1404 USA18Minnesota Institute for Astrophysics University of Minnesota Minneapolis MN 55455 USA

Accepted 2018 July 1 Received 2018 June 8 in original form 2018 April 17

ABSTRACT

We present a detailed study of the 2016 eruption of nova V407 Lupi (ASASSN-16kt)including optical near-infrared X-ray and ultraviolet data from SALT SMARTSSOAR Chandra Swift and XMM-Newton Timing analysis of the multiwavelengthlight-curves shows that from 168 days post-eruption and for the duration of the X-ray supersoft source phase two periods at 565 s and 357 h are detected We suggestthat these are the rotational period of the white dwarf and the orbital period ofthe binary respectively and that the system is likely to be an intermediate polarThe optical light-curve decline was very fast (t2 6 29 d) suggesting that the whitedwarf is likely massive (amp 125M⊙) The optical spectra obtained during the X-raysupersoft source phase exhibit narrow complex and moving emission lines of He iialso characteristics of magnetic cataclysmic variables The optical and X-ray datashow evidence for accretion resumption while the X-ray supersoft source is still onpossibly extending its duration

Key words stars individual (V407 Lup) ndash novae cataclysmic variables ndash whitedwarfs

arX

iv1

807

0070

6v1

[as

tro-

phS

R]

2 J

ul 2

018

2 Aydi et al

1 INTRODUCTION

Classical novae (CNe) are stellar eruptions that take placewithin the surface layers of accreting white dwarfs (WDs)in cataclysmic variable (CV) systems These are interactingbinaries consisting of a WD primary accreting from a sec-ondary that typically fills its Roche lobe (see eg Warner1995) The hydrogen-rich material accreted by the WD ac-cumulates on its surface causing an increase in pressure anddensity to a level where a thermonuclear runaway (TNR) istriggered (Starrfield 1989 Starrfield Iliadis amp Hix 2008 Joseamp Shore 2008 Starrfield Iliadis amp Hix 2016) Due to thisa part of the accreted envelope is ejected The fast expand-ing envelope moves together with the optical photosphere(Hachisu amp Kato 2006 2014) in what is known as the ldquofire-ball stagerdquo During this the brightness of the star increasesby 8 up to 15 mag (Payne-Gaposchkin 1964) and reaches itsmaximum visual brightness when the optical photosphereattains its maximum radius (Warner 1995 Hachisu amp Kato2014) Following the TNR the remaining hydrogen-rich ac-creted envelope continuously burns on the WD surface andmay become visible when the expanding ejecta become op-tically thin

After maximum the optical depth of the expandingejecta decreases progressively and therefore the optical pho-tosphere starts shrinking This shifts the emission to higherenergies and would peak in the soft X-ray band and a Su-perSoft Source (SSS) emerges unless the ejecta become op-tically thin before that happens in which case the SSS emer-gence is due to the visibility of the WD photosphere whichis at that time extended due to the ongoing nuclear burning(Krautter 2008 Osborne 2015) As the hydrogen in the sur-face layers is consumed the SSS ldquoturns offrdquo and the systemeventually evolves back to quiescence (for a series of reviewsof CNe see Bode amp Evans 2008)

Observing CNe across many wavelengths is essential toprovide a full picture of the eruption and its characteris-tics such as the different physical parameters (eg tem-peratures and velocities) the gas ejection mechanisms theradiative processes the nuclear reactions the shocks in theejectawind the dust formation and the properties of theprogenitor However despite their importance systematicmultiwavelength studies are still lacking and only a few CNehave been followed panchromatically and in detail (see egShore et al 2013 Schwarz et al 2015 Bode et al 2016 Shoreet al 2016 Darnley et al 2016 Li et al 2017 Mason et al2018 Aydi et al 2018) Therefore observing these eventsin multiple frequency bands is essential for a better under-standing of these individual events and ultimately adding tothe wealth of knowledge in the field

While most nova eruptions occur in non-magnetic CVswhere the WD is usually weakly magnetized (magnetic fieldof the WD is 106 G) a few novae have been seen to occurin magnetic CVs (mCVs) These systems form a sub-groupof CVs in which the WD is highly magnetized and they in-clude two main sub-types polars which have the strongestmagnetic fields (amp 107 G) and intermediate polars (IPs) withweaker fields (106 ndash 107 G) The strong magnetic field of po-lars causes the spin (rotational) period of the WD and theorbital period of the binary to synchronize This is not thecase for IPs where the optical X-ray and ultraviolet (UV)light-curves might show multiple periodicities modulated

on the orbital period of the binary (with typical values of3 to 6 h) the spin period of the WD (with values rangingbetween sim 05 up to sim 70min) and the sideband periods(typically dominated by the beat period) Although in suchsystems the mass-accretion mechanism is dominated at somepoint by the magnetic field of the WD the accretion ontothe surface of WD can still result in a nova eruption Fora comprehensive review of mCVs and their mass-transfer-accretion mechanisms see Warner (1995) and Hellier (2001)

The only nova known to have erupted in a polar sys-tem is V1500 Cyg (Kaluzny amp Semeniuk 1987 StockmanSchmidt amp Lamb 1988) The eruption which occurred in1975 was one of the most luminous on record along with CPPup (Warner 1985) and nova SMCN 2016-10a (Aydi et al2018) On the other hand a few novae have been either con-firmed (GK Per DQ Her V4743 Sgr and Nova Scorpii 1437AD) or suggested (V533 Her DD Cir V1425 Aql M31N2007-12b and V2491 Cyg) to occur in IPs (see eg Os-borne et al 2001 King Osborne amp Schenker 2002 WarnerampWoudt 2002 Bianchini et al 2003 Woudt ampWarner 20032004 Leibowitz et al 2006 Dobrotka amp Ness 2010 Pietschet al 2011 Zemko et al 2016 Potter amp Buckley 2018 andreferences therein) The effect of the magnetic field on thenova eruption and progress is not well understood

In this paper we present a multiwavelength studyof nova V407 Lupi which was discovered by the All-SkyAutomated Survey for Supernovae (ASAS-SN)1 on HJD24576555 (2016 September 240 UT Stanek et al 2016)at V = 91 and is located at equatorial coordinates of(α δ)J20000 = (15h29m0182s ndash4449prime40primeprime89) and Galacticcoordinates of (l b) = (33009 9573) The last ASAS-SNpre-discovery observation reported the source at V gt 155on HJD 24576515 Therefore in the following we assumeHJD 24576555 (2016 September 240 UT) is t0 (eruptionstart) Our study consists of a set of comprehensive multi-wavelength observations obtained with the aim of studyingin detail the post-eruption behaviour of the nova at opti-cal near-infrared (NIR) ultraviolet (UV) and X-ray wave-lengths Note that this object has not been seen in eruptionbefore This paper is structured as follows in Section 2 wepresent the observations and data reduction The analysisof the photometric and spectroscopic results are given inSections 3 and 4 respectively We present the discussion inSection 5 while Section 6 contains a summary and the con-clusions

2 OBSERVATIONS AND DATA REDUCTION

21 SMARTS observations and data reduction

Since 2016 September 26 (day 2) the eruption has beenmonitored using the Small and Moderate Aperture ResearchTelescope System (SMARTS) to obtain optical BVRI andNIR JHK photometric observations The integration timesat JHK were 15 s (three 5 s dithered images) before 2016October 6 (day 12) and 30 s thereafter Optical observationsare single images of 30 s 25 s 20 s and 20 s integrations re-spectively in BVRI prior to 2016 October 6 (day 12) and

1 httpwwwastronomyohio-stateeduasassnindexshtml

nova V407 Lup 3

uniformly 50 s thereafter The procedure followed for reduc-ing the SMARTS photometry is detailed in Walter et al(2012) Fig B1 represents the SMARTS BVRI and JHK

observations See Table D1 for a log of the observations

22 SALT high-resolution Echelle spectroscopy

We used the SALT High Resolution Spectrograph (HRSBarnes et al 2008 Bramall et al 2010 Bramall et al 2012Crause et al 2014) to obtain observations on the nights of2017 March 6 April 20 May 31 June 13 29 July 05 24 29and August 06 (respectively days 164 209 250 263 279285 304 309 and 317) HRS a dual beam fibre-fed Echellespectrograph housed in a temperature stabilized vacuumtank was used in the low-resolution (LR) mode to obtainall the observations This mode provides two spectral rangesblue (3800 ndash 5550 A) and red (5450 ndash 9000 A) at a resolutionof R sim 15000 A weekly set of HRS calibrations includingfour ThAr + Ar arc spectra and four spectral flats is ob-tained in all the modes (low medium and high-resolution)All the HRS observations are 1800 s exposures See Table D2for a log of the observations

The primary reduction was conducted using the SALTscience pipeline (Crawford et al 2010) which includes over-scan correction bias subtraction and gain correction Therest of the reduction and relative flux calibration wasdone using the standard MIDAS FEROS (Stahl Kaufer ampTubbesing 1999) and echelle (Ballester 1992) packages Thereduction procedure is described in detail in Kniazev Gvara-madze amp Berdnikov (2016) Note that absolute flux calibra-tion is not feasible with SALT data As part of the SALTdesign the effective area of the telescope constantly changeswith the moving pupil during the track and exposures

23 SOAR medium-resolution spectroscopy

We performed optical spectroscopy of the nova on 2017 June20 (day 270) using the Goodman spectrograph (ClemensCrain amp Anderson 2004) on the 41m Southern Astrophys-ical Research (SOAR) telescope We obtained a single ex-posure of 600 s using a 600 lmmminus1 grating and a 107primeprime slitto provide a resolution of sim 13 A over the range of 3200 ndash6800 A The spectrum was reduced using the standard rou-tine and a relative flux calibration has been applied

We also obtained two spectra using the Goodman spec-trograph on the nights of 2018 January 20 and 21 (days484 and 485) each of 20min exposure For these two ob-servations we used a setup with a 2100 lmmminus1 grating anda 095primeprimeslit yielding a resolution of about 088 A full widthat half maximum (FWHM) (56 km sminus1) over a wavelengthrange of 4280 ndash 4950 AA The spectra were reduced and op-timally extracted in the usual manner

24 Swift X-ray and UV observations

241 Swift XRT observations

The Neil Gehrels Swift Observatory (hereafter Swift Gehrels et al 2004) first observed V407 Lup on 2016 Septem-ber 26 two days after the discovery The high optical bright-ness of the nova at this time meant that the UVOpticalTelescope (UVOT Roming et al 2005) could not be used

and the X-ray Telescope (XRT Burrows et al 2005) neededto be operated in Windowed Timing (WT) mode for theinitial observation in order to minimize the effects of op-tical loading whereby a large number of optical photonscan pile-up to appear as a false X-ray signature2 No X-raysource was detected at this time (lt 002 count sminus1 grade 0ndash single pixel events) or in the individual subsequent dailyobservations taken in Photon Counting (PC) mode between2016 October 2 and 9 (days 8 ndash 15) Coadding these PC datathere is a source detected at the 99 per cent confidence levelwith a count rate of (14+05

minus04) times10minus3 count sminus1 (grade 0)There was also a suggestion of a detection in the October10 (day 16) data alone These detections could not be con-firmed at a higher significance before V407 Lup became tooclose to the Sun for Swift to observe on 2016 October 13(day 19)

Observations recommenced on 2017 February 21 (day150) finding a bright supersoft X-ray source (Beardmoreet al 2017) with a count rate of 561plusmn 03 count sminus1 Bythis time the UVOT could also be safely operated and a UVsource with uvw2 = 1349plusmn 002mag was measured Dailyobservations were performed between 2017 February 22 and28 (days 151 ndash 157) followed by observations approximatelyevery two to three days until 2017 April 24 (day 212) Obser-vations were continued though with a slightly lower cadenceof every four days until 2017 September 10 (day 351) withthe cadence then decreasing to every eight days until 2017October 13 (day 384) when the Sun constraint began againA final dataset was obtained once the source again emergedfrom the Sun constraint on 2018 January 23 (day 486) Allobservations were typically 05-1 ks in duration with the ex-ception of the final observation which was 3 ks In addition tothis regular monitoring a high cadence campaign was per-formed between 2017 July 4 ndash 8 (days 283 ndash 287) aimed atpinning down the periodicity seen in the UVOT light-curve(see Section 34) In this case sim 500 s snapshots were takenapproximately every 6 h Because of complications causedby a nearby bright star no UVOT data were actually col-lected and so the campaign was repeated from 2017 July28 ndashAugust 1 (days 307 ndash 311) using an offset pointing toavoid this problem The Swift XRT observations log is givenin Table D3

The Swift data were processed with the standardHEASoft tools (version 620)3 and analysed using themost up-to-date calibration files All observations be-tween 2017 February 21 and August 11 (days 150 and321 post-eruption) inclusive were obtained using WTmode because of the high brightness of the X-raysource These data were extracted using a circular re-gion of a radius of 20 pixels (1 pixel = 236 arcsec)for the source and a background annulus as describedat httpwwwswiftacukanalysisxrtbackscalphp Ob-servations from 2017 August 15 (day 325) onwards weretaken with the XRT in PC mode the first three of thesesuffered from pile-up so an annulus (outer radius 30 pix-els inner exclusion radius decreasing from seven to three

2 httpwwwswiftacukanalysisxrtoptical_loading

php3 httpsheasarcgsfcnasagovFTPsoftwareftools

releasearchiveRelease_Notes_620

4 Aydi et al

pixels) was used when extracting the source counts Thelater data were analysed using circular regions decreasingin radius from 20 to 10 pixels as the source further fadedBackground counts were estimated from near-by source-freecircular regions of 60 pixels radius

The X-ray spectra were binned to provide a minimum of1 count binminus1 to facilitate Cash statistic (Cash 1979) fittingwithin xspec (Arnaud 1996)

242 Swift UVOT observations

Observations with the Swift UVOT instrument startedonce the nova could be observed again after its passage be-hind the Sun On 2017 February 26 (day 156) the nova wasobserved in the uvw2 (λcentral = 1928 A) filter the next dayin the uvm2 (λcentral = 2246 A) filter Photometric observa-tions continued until 2017 October 13 (day 385) A log ofthe observations is given in Table D3

Weekly observations with the UV grism started on 2017March 7 until April 23 (day 165 until day 212) when theywere discontinued as the brightness was too low The UVgrism provides a spectral range of sim 1700 ndash 5100 A Thelog of the observations are given in Table D4 The eightgrism observations were reduced using the UVOTPY soft-ware (Kuin 2014) using the calibration described in Kuinet al (2015) with a recent update to the sensitivity affect-ing the response of the UV grism mainly below 2000 A 4The net continuum counts in the UV part of the individualspectra are very low so the first four and last four spectrawere summed which also improved the signal to noise (SN)of the weaker lines significantly No reddening correction hasbeen applied to the spectra

25 XMM-Newton X-ray and UV observations

V407 Lup was observed by XMM-Newton from 2017 March11 1145 to 1708 UT 1685 days post-eruption with anexposure duration of 23 ks (Ness et al 2017) The XMM-

Newton observatory consists of five different X-ray in-struments behind three mirrors plus an optical monitor(OMMason et al 2001 Talavera 2009) which all observesimultaneously For a full description of the X-ray instru-ments onboard of XMM-Newton see Jansen et al (2001)den Herder et al (2001) Struder et al (2001) Turner et al(2001) and Aschenbach (2002)

In this work we only used the spectra and light-curvesfrom the Reflection Grating Spectrometer (RGSden Herderet al 2001) the light-curves from the European PhotonImaging Camera (EPIC)pn (Struder et al 2001) and theOM light-curvesgrism spectra (see Table D5 for a log of theobservations)

The OM took five exposures one of which with the vis-ible grism provided a spectrum between 3000 and 6000 A(XMM-Newton could not take a U -grism spectrum becauseof a contaminating nearby star so we only obtained a V -grism spectrum) and four with Science User Defined imag-ing plus fast mode which provided a UV light-curve withthe uvw1 (λ sim 2910plusmn 830 A) filter

4 seehttpmssluclacuknpmkGrism

The RGS provides a spectral range of 6 ndash 38 A fully cov-ering the Wien tail of SSS spectra (30 ndash 80 eV) while theRayleigh Jeans tail is not visible (owing to interstellar ab-sorption) It also provides X-ray light-curves in the sameenergy range RGS consists of two instruments RGS1 andRGS2 One of the nine CCDs in RGS2 is suffering from sometechnical problems We Combine the RGS1 and RGS2 spec-tra in order to overcome the problem of the broken CCD inthe RGS2 instrument The EPICpn instrument was oper-ated in Timing Mode with medium filter providing anotherX-ray light-curve

26 Chandra X-ray observations

V407 Lup was observed with the Chandra X-ray Observa-

tory (hereafter Chandra Weisskopf et al 2000) using theHigh Resolution Camera (HRC Murray et al 1997) andthe Low Energy Transmission Grating (LETG Brinkmanet al 2000ab) on 2018 August 30 (day 3406) with an ex-posure time of 34 ks The LETG instrument provides a spec-trum over a range of sim 12 ndash 175 A (007 ndash 1033 keV) how-ever most of the flux from nova V407 Lup is in the range15 ndash 50 A (024 ndash 082 keV) The light-curve provided by theHRC instrument has a range of 006 ndash 10 keV

3 PHOTOMETRIC RESULTS AND ANALYSIS

31 Optical light-curve parameters

Several parameters characterize nova light-curves includingthe rise rate the rise time to maximum light the maximumlight the decline rate and the decline behaviour (see egHounsell et al 2010 Cao et al 2012) Nova V407 Lup wasnot extensively observed during its rise to maximum

Based on V -band and visual (Vis) measurements fromSMARTS and the American Association of Variable StarObservers (AAVSO)5 we see that the nova reached V =68 04 d after t0 Then it was reported at V = 64 on HJD245765624 (Prieto 2016) and Vis = 63 on HJD 245765657reaching Vis = 56 on HJD 245765690 (see Table C1) Halfa day later the nova was at V = 633 and Vis = 65 thendropped to V = 785 in around two days indicating thatthe decline had started Hence we assume HJD 245765690as tmax (day 14)

Caution is required in interpreting magnitudes of no-vae This is particularly so for Vis where Hα may con-tribute to the flux detected by eye but not to CCD V -magnitudes Furthermore normal transformation relationsfor CCD magnitudes cannot be used for objects with strongemission lines Unfortunately no spectra were taken aroundmaximum light so the contribution and development of lineemission is unknown until later times (day 5 Izzo et al2018) In the following we assume that maximum light wasVmax 6 60 on HJD 245765690 Since novae do not usu-ally have strong line emission at maximum light (van denBergh amp Younger 1987) and given that simultaneous V andVis measurements show very similar magnitudes before andafter maximum light it may be that V sim Vis =56 onHJD 245765690 is a reasonable estimate However we use

5 httpswwwaavsoorg

nova V407 Lup 5

Figure 1 V -band light-curve using data from SMARTS (green circles) AAVSO (red squares) and ATels from Stanek et al (2016)(yellow circle) Chen et al (2016) (blue stars) and Prieto (2016) (cyan diamond) A zoom-in plot around the first sim 25 days is added forclarity Within this plot the blue dashed line represents a power-law fit to the data (before day 25) excluding the discovery measurementthe black horizontal dashed lines represent from top to bottom Vmax Vmax + 2 and Vmax + 3 respectively the black vertical dashedlines represent from left to right tmax t2 and t3 respectively

Vmax 6 60 for the rest of the analysis to avoid overestimat-ing the maximum brightness

We construct a V -band light-curve by combining thepublished photometry from different telescopes and instru-ments and the SMARTS data (see Fig 1) noting the abovecaveat This shows a rapid rise and a sharp peak followed bya fast decline Such behaviour is characteristic of the S-classnova light-curves (Strope Schaefer amp Henden 2010) whichis considered stereotypical for a nova However the broad-band light-curves show day-to day variability after t3 similarto that of an O-class light-curves (see Fig B1) It is worthnoting that there is no presence for a plateau in the tail ofthe light-curve (after sim 3 ndash 6mag from maximum) similar tothat seen in recurrent novae (see Schaefer 2010)

Although there is a gap in the light-curve between day26 and day 107 (due to solar constraints) we fit a power-lawto the light-curve (only early decline - before day 25) andderive t2 6 29plusmn 05 d t3 6 68 plusmn 10 d and a power index ofsim 0166 We took into account the use of measurements fromdifferent instruments by increasing (times 2) the uncertainty ont2 and t3

With t2 6 29 d and a decline rate of sim 069mag dminus1

over t2 nova V407 Lup is a ldquovery-fastrdquo nova in the classifi-

6 Izzo et al (2018) have derived longer values of t2 and t3 This

was due to misinterpretation of their light-curve (private commu-nication with L Izzo)

Figure 2 The UVOT uvw2 (λcentral = 1928 A) light-curve plot-ted against day since t0

cation of Payne-Gaposchkin (1964) and is one of the fastestknown examples Only a few other novae have shown adecline time t2 30 days including M31N 2008-12a USco V1500 Cyg V838 Her V394 CrA and V4160 Sgr (seeegYoung et al 1976 Schaefer 2010 Munari et al 2011and table 5 in Darnley et al 2016)

6 Aydi et al

Figure 3 First panel the XMM-Newton EPICpn X-ray light-curve Second panel the RGS1 X-ray light-curve The X-ray light-curvesshow two broad (sim 2 ndash 3 ks wide) dips separated by 126 ks consistent with the 357 h modulation seen in the Swift-UVOT data (seeSection 34) Third panel the uvw1 OM fast mode (black) and imaging mode (red) light-curves The first OM exposure is a grismspectrum hence the light-curves only start at sim 5 ks The blue line represents a best fit sine curve to the OM data with a period of376plusmn03 h This period is consistent within uncertainty with the period seen in the Swift-UVOT data (see Section 34 for further

discussion on the timing analysis) The y-axis in all the panels are counts per second

32 The Swift UVOT light-curve

The UVOT uvw2 light-curve (Fig 2) declined slowly afterday 150 before flattening off at around a uvw2 magnitude of142 by around day 200 with signs of a 01mag variabilitysuperimposed on the overall decline We find periodicities inthe light-curve which we discuss in Section 34 The UV datarebrightened after day 300 to reach a uvw2 magnitude of137 around day 320 Then the brightness started to declineagain reaching a uvw2 magnitude of 149 at day sim 385Fig B2 shows a direct comparison between the Swift X-raySwift UV and SMARTS optical light-curves

33 The XMM-Newton and Chandra UV and

X-ray light-curves

The XMM-Newton RGS1 X-ray light-curve the EPICpnX-ray light-curve and the OM UV fast and imaging modeslight-curves are all presented in Fig 3 The XMM-Newton

RGS1 light-curve shows evidence of variability on differenttimescales This includes two broad (sim 2 ndash 3 ks wide) dipsseparated by sim 126 ks sim 35 h The OM fast mode UV light-curve shows a variation characterized by a 376plusmn03 h periodwhich is also consistent (within uncertainty) with the dura-tion between the dips in the RGS1 light-curve Howeverno other periodicity is found in the OM light-curve TheChandra HRC light-curve (Fig 4) also shows evidence forshort-term variability which we discuss in Section 34

Figure 4 The Chandra High Resolution Camera (HRC) X-ray

light-curve The HRC energy range is 006 ndash 10 keV There is aclear short-term variability in the light-curve and it is discussed

in Section 34

34 Timing analysis

Since the SwiftUVOT uvw2 light-curve (Fig 2) shows evi-dence for intrinsic variability superimposed on the long termdecline we searched for periodicities in the data Fig 5shows a Lomb-Scargle periodogram (LSP) of the barycen-tric corrected uvw2 light-curve after subtracting a third or-der polynomial to remove the long term trend The mostsignificant peaks occur at periods of 35731 plusmn 00014 h and

nova V407 Lup 7

Figure 5 Upper panel the Lomb Scargle periodogram of thedetrended uvw2 light-curve revealing significant modulation atfrequencies of 67168 dminus1 (777times10minus5 sminus1 P = 3573 h) and217696 dminus1 (252 times 10minus4 sminus1 P = 1102 h with estimated un-certainties of 00003 dminus1) marked with red dashed lines Lowerpanel periodogram of the observing window showing Swift rsquos or-

bital frequency (dotted line)

11024plusmn 00002 h which are aliases of each other caused bySwift rsquos 16 h orbit One of these periods might represent theorbital period of the binary (Porb) The low cadence of theUVOT data means we cannot break the degeneracy betweenthe periods (the high-cadence observation interval betweendays 307 and 311 failed to resolve this issue) The amplitudeof the modulation is approximately 01mag

We performed a similar Lomb-Scargle analysis of theXMM-Newton RGS1 and Chandra HRC data (see Fig 6)The light-curve of the former shows considerable flux vari-ations while the flux in the latter is approximately con-stant over the duration of the observation The XMM-

Newton periodogram is dominated by a low frequency ofΩ = 777times10minus5 sminus1 (357 h) This signal only covers sim twocycles of the 357 h period and cannot be believed in iso-lation However as this signal is consistent with the longperiod seen in the UVOT data we identify it as the same357 h modulation

The long-timescale modulation and variability seen inthe XMM-Newton light-curves the time between the dipsand the weak modulation observed in the OM light-curve(see Figs 3 and 6) are all consistent with the 357 h signalseen in the Swift UV light-curve In addition the absenceof any variability in the XMM-Newton RGS and OM light-curves related to the 11 h signal (seen in the Swift UV light-curve) has the potential to rule out the modulation at thisshort period Therefore we assume that the 357 h is mostlikely the Porb of the binary

Ignoring this long-term variability seen in the low fre-quency part of XMM-Newton periodogram we find signifi-cant signals at ω ≃ 177times10minus3 sminus1 in both (XMM-Newton

and Chandra) light-curves More precisely the periodicitieswe found are at 56504plusmn 068 s for the Chandra HRC light-curve and 56496plusmn 050 s for the XMM-Newton RGS light-curve The SwiftXRT observations are almost all too shortto usefully constrain the modulation at this short period

Figure 6 Lomb Scargle periodograms of the XMM-NewtonRGS1 (top) and the Chandra HRC (bottom) light-curves Vari-ous frequencies are indicated where Ω = 1Porb and ω = 1Pspin

(see text for more details) We present zoom-in plots around the177times10minus3 sminus1 in each periodogram

Figure 7 Top from top to bottom soft RGS1 light-curve hard

RGS1 light-curve and the hardness ratio The energy bandsfor the soft and hard light-curves are 235 ndash 370 A (03325 ndash

0528 keV) and 150 ndash 235 A (0528 ndash 0827 keV) respectively Bot-tom same as top but all folded at the 565 s spin period andnormalized to their respective mean rates Therefore the y-axisrepresents the relative amplitude of the modulation in both thesoft and hard bands

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 4: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

2 Aydi et al

1 INTRODUCTION

Classical novae (CNe) are stellar eruptions that take placewithin the surface layers of accreting white dwarfs (WDs)in cataclysmic variable (CV) systems These are interactingbinaries consisting of a WD primary accreting from a sec-ondary that typically fills its Roche lobe (see eg Warner1995) The hydrogen-rich material accreted by the WD ac-cumulates on its surface causing an increase in pressure anddensity to a level where a thermonuclear runaway (TNR) istriggered (Starrfield 1989 Starrfield Iliadis amp Hix 2008 Joseamp Shore 2008 Starrfield Iliadis amp Hix 2016) Due to thisa part of the accreted envelope is ejected The fast expand-ing envelope moves together with the optical photosphere(Hachisu amp Kato 2006 2014) in what is known as the ldquofire-ball stagerdquo During this the brightness of the star increasesby 8 up to 15 mag (Payne-Gaposchkin 1964) and reaches itsmaximum visual brightness when the optical photosphereattains its maximum radius (Warner 1995 Hachisu amp Kato2014) Following the TNR the remaining hydrogen-rich ac-creted envelope continuously burns on the WD surface andmay become visible when the expanding ejecta become op-tically thin

After maximum the optical depth of the expandingejecta decreases progressively and therefore the optical pho-tosphere starts shrinking This shifts the emission to higherenergies and would peak in the soft X-ray band and a Su-perSoft Source (SSS) emerges unless the ejecta become op-tically thin before that happens in which case the SSS emer-gence is due to the visibility of the WD photosphere whichis at that time extended due to the ongoing nuclear burning(Krautter 2008 Osborne 2015) As the hydrogen in the sur-face layers is consumed the SSS ldquoturns offrdquo and the systemeventually evolves back to quiescence (for a series of reviewsof CNe see Bode amp Evans 2008)

Observing CNe across many wavelengths is essential toprovide a full picture of the eruption and its characteris-tics such as the different physical parameters (eg tem-peratures and velocities) the gas ejection mechanisms theradiative processes the nuclear reactions the shocks in theejectawind the dust formation and the properties of theprogenitor However despite their importance systematicmultiwavelength studies are still lacking and only a few CNehave been followed panchromatically and in detail (see egShore et al 2013 Schwarz et al 2015 Bode et al 2016 Shoreet al 2016 Darnley et al 2016 Li et al 2017 Mason et al2018 Aydi et al 2018) Therefore observing these eventsin multiple frequency bands is essential for a better under-standing of these individual events and ultimately adding tothe wealth of knowledge in the field

While most nova eruptions occur in non-magnetic CVswhere the WD is usually weakly magnetized (magnetic fieldof the WD is 106 G) a few novae have been seen to occurin magnetic CVs (mCVs) These systems form a sub-groupof CVs in which the WD is highly magnetized and they in-clude two main sub-types polars which have the strongestmagnetic fields (amp 107 G) and intermediate polars (IPs) withweaker fields (106 ndash 107 G) The strong magnetic field of po-lars causes the spin (rotational) period of the WD and theorbital period of the binary to synchronize This is not thecase for IPs where the optical X-ray and ultraviolet (UV)light-curves might show multiple periodicities modulated

on the orbital period of the binary (with typical values of3 to 6 h) the spin period of the WD (with values rangingbetween sim 05 up to sim 70min) and the sideband periods(typically dominated by the beat period) Although in suchsystems the mass-accretion mechanism is dominated at somepoint by the magnetic field of the WD the accretion ontothe surface of WD can still result in a nova eruption Fora comprehensive review of mCVs and their mass-transfer-accretion mechanisms see Warner (1995) and Hellier (2001)

The only nova known to have erupted in a polar sys-tem is V1500 Cyg (Kaluzny amp Semeniuk 1987 StockmanSchmidt amp Lamb 1988) The eruption which occurred in1975 was one of the most luminous on record along with CPPup (Warner 1985) and nova SMCN 2016-10a (Aydi et al2018) On the other hand a few novae have been either con-firmed (GK Per DQ Her V4743 Sgr and Nova Scorpii 1437AD) or suggested (V533 Her DD Cir V1425 Aql M31N2007-12b and V2491 Cyg) to occur in IPs (see eg Os-borne et al 2001 King Osborne amp Schenker 2002 WarnerampWoudt 2002 Bianchini et al 2003 Woudt ampWarner 20032004 Leibowitz et al 2006 Dobrotka amp Ness 2010 Pietschet al 2011 Zemko et al 2016 Potter amp Buckley 2018 andreferences therein) The effect of the magnetic field on thenova eruption and progress is not well understood

In this paper we present a multiwavelength studyof nova V407 Lupi which was discovered by the All-SkyAutomated Survey for Supernovae (ASAS-SN)1 on HJD24576555 (2016 September 240 UT Stanek et al 2016)at V = 91 and is located at equatorial coordinates of(α δ)J20000 = (15h29m0182s ndash4449prime40primeprime89) and Galacticcoordinates of (l b) = (33009 9573) The last ASAS-SNpre-discovery observation reported the source at V gt 155on HJD 24576515 Therefore in the following we assumeHJD 24576555 (2016 September 240 UT) is t0 (eruptionstart) Our study consists of a set of comprehensive multi-wavelength observations obtained with the aim of studyingin detail the post-eruption behaviour of the nova at opti-cal near-infrared (NIR) ultraviolet (UV) and X-ray wave-lengths Note that this object has not been seen in eruptionbefore This paper is structured as follows in Section 2 wepresent the observations and data reduction The analysisof the photometric and spectroscopic results are given inSections 3 and 4 respectively We present the discussion inSection 5 while Section 6 contains a summary and the con-clusions

2 OBSERVATIONS AND DATA REDUCTION

21 SMARTS observations and data reduction

Since 2016 September 26 (day 2) the eruption has beenmonitored using the Small and Moderate Aperture ResearchTelescope System (SMARTS) to obtain optical BVRI andNIR JHK photometric observations The integration timesat JHK were 15 s (three 5 s dithered images) before 2016October 6 (day 12) and 30 s thereafter Optical observationsare single images of 30 s 25 s 20 s and 20 s integrations re-spectively in BVRI prior to 2016 October 6 (day 12) and

1 httpwwwastronomyohio-stateeduasassnindexshtml

nova V407 Lup 3

uniformly 50 s thereafter The procedure followed for reduc-ing the SMARTS photometry is detailed in Walter et al(2012) Fig B1 represents the SMARTS BVRI and JHK

observations See Table D1 for a log of the observations

22 SALT high-resolution Echelle spectroscopy

We used the SALT High Resolution Spectrograph (HRSBarnes et al 2008 Bramall et al 2010 Bramall et al 2012Crause et al 2014) to obtain observations on the nights of2017 March 6 April 20 May 31 June 13 29 July 05 24 29and August 06 (respectively days 164 209 250 263 279285 304 309 and 317) HRS a dual beam fibre-fed Echellespectrograph housed in a temperature stabilized vacuumtank was used in the low-resolution (LR) mode to obtainall the observations This mode provides two spectral rangesblue (3800 ndash 5550 A) and red (5450 ndash 9000 A) at a resolutionof R sim 15000 A weekly set of HRS calibrations includingfour ThAr + Ar arc spectra and four spectral flats is ob-tained in all the modes (low medium and high-resolution)All the HRS observations are 1800 s exposures See Table D2for a log of the observations

The primary reduction was conducted using the SALTscience pipeline (Crawford et al 2010) which includes over-scan correction bias subtraction and gain correction Therest of the reduction and relative flux calibration wasdone using the standard MIDAS FEROS (Stahl Kaufer ampTubbesing 1999) and echelle (Ballester 1992) packages Thereduction procedure is described in detail in Kniazev Gvara-madze amp Berdnikov (2016) Note that absolute flux calibra-tion is not feasible with SALT data As part of the SALTdesign the effective area of the telescope constantly changeswith the moving pupil during the track and exposures

23 SOAR medium-resolution spectroscopy

We performed optical spectroscopy of the nova on 2017 June20 (day 270) using the Goodman spectrograph (ClemensCrain amp Anderson 2004) on the 41m Southern Astrophys-ical Research (SOAR) telescope We obtained a single ex-posure of 600 s using a 600 lmmminus1 grating and a 107primeprime slitto provide a resolution of sim 13 A over the range of 3200 ndash6800 A The spectrum was reduced using the standard rou-tine and a relative flux calibration has been applied

We also obtained two spectra using the Goodman spec-trograph on the nights of 2018 January 20 and 21 (days484 and 485) each of 20min exposure For these two ob-servations we used a setup with a 2100 lmmminus1 grating anda 095primeprimeslit yielding a resolution of about 088 A full widthat half maximum (FWHM) (56 km sminus1) over a wavelengthrange of 4280 ndash 4950 AA The spectra were reduced and op-timally extracted in the usual manner

24 Swift X-ray and UV observations

241 Swift XRT observations

The Neil Gehrels Swift Observatory (hereafter Swift Gehrels et al 2004) first observed V407 Lup on 2016 Septem-ber 26 two days after the discovery The high optical bright-ness of the nova at this time meant that the UVOpticalTelescope (UVOT Roming et al 2005) could not be used

and the X-ray Telescope (XRT Burrows et al 2005) neededto be operated in Windowed Timing (WT) mode for theinitial observation in order to minimize the effects of op-tical loading whereby a large number of optical photonscan pile-up to appear as a false X-ray signature2 No X-raysource was detected at this time (lt 002 count sminus1 grade 0ndash single pixel events) or in the individual subsequent dailyobservations taken in Photon Counting (PC) mode between2016 October 2 and 9 (days 8 ndash 15) Coadding these PC datathere is a source detected at the 99 per cent confidence levelwith a count rate of (14+05

minus04) times10minus3 count sminus1 (grade 0)There was also a suggestion of a detection in the October10 (day 16) data alone These detections could not be con-firmed at a higher significance before V407 Lup became tooclose to the Sun for Swift to observe on 2016 October 13(day 19)

Observations recommenced on 2017 February 21 (day150) finding a bright supersoft X-ray source (Beardmoreet al 2017) with a count rate of 561plusmn 03 count sminus1 Bythis time the UVOT could also be safely operated and a UVsource with uvw2 = 1349plusmn 002mag was measured Dailyobservations were performed between 2017 February 22 and28 (days 151 ndash 157) followed by observations approximatelyevery two to three days until 2017 April 24 (day 212) Obser-vations were continued though with a slightly lower cadenceof every four days until 2017 September 10 (day 351) withthe cadence then decreasing to every eight days until 2017October 13 (day 384) when the Sun constraint began againA final dataset was obtained once the source again emergedfrom the Sun constraint on 2018 January 23 (day 486) Allobservations were typically 05-1 ks in duration with the ex-ception of the final observation which was 3 ks In addition tothis regular monitoring a high cadence campaign was per-formed between 2017 July 4 ndash 8 (days 283 ndash 287) aimed atpinning down the periodicity seen in the UVOT light-curve(see Section 34) In this case sim 500 s snapshots were takenapproximately every 6 h Because of complications causedby a nearby bright star no UVOT data were actually col-lected and so the campaign was repeated from 2017 July28 ndashAugust 1 (days 307 ndash 311) using an offset pointing toavoid this problem The Swift XRT observations log is givenin Table D3

The Swift data were processed with the standardHEASoft tools (version 620)3 and analysed using themost up-to-date calibration files All observations be-tween 2017 February 21 and August 11 (days 150 and321 post-eruption) inclusive were obtained using WTmode because of the high brightness of the X-raysource These data were extracted using a circular re-gion of a radius of 20 pixels (1 pixel = 236 arcsec)for the source and a background annulus as describedat httpwwwswiftacukanalysisxrtbackscalphp Ob-servations from 2017 August 15 (day 325) onwards weretaken with the XRT in PC mode the first three of thesesuffered from pile-up so an annulus (outer radius 30 pix-els inner exclusion radius decreasing from seven to three

2 httpwwwswiftacukanalysisxrtoptical_loading

php3 httpsheasarcgsfcnasagovFTPsoftwareftools

releasearchiveRelease_Notes_620

4 Aydi et al

pixels) was used when extracting the source counts Thelater data were analysed using circular regions decreasingin radius from 20 to 10 pixels as the source further fadedBackground counts were estimated from near-by source-freecircular regions of 60 pixels radius

The X-ray spectra were binned to provide a minimum of1 count binminus1 to facilitate Cash statistic (Cash 1979) fittingwithin xspec (Arnaud 1996)

242 Swift UVOT observations

Observations with the Swift UVOT instrument startedonce the nova could be observed again after its passage be-hind the Sun On 2017 February 26 (day 156) the nova wasobserved in the uvw2 (λcentral = 1928 A) filter the next dayin the uvm2 (λcentral = 2246 A) filter Photometric observa-tions continued until 2017 October 13 (day 385) A log ofthe observations is given in Table D3

Weekly observations with the UV grism started on 2017March 7 until April 23 (day 165 until day 212) when theywere discontinued as the brightness was too low The UVgrism provides a spectral range of sim 1700 ndash 5100 A Thelog of the observations are given in Table D4 The eightgrism observations were reduced using the UVOTPY soft-ware (Kuin 2014) using the calibration described in Kuinet al (2015) with a recent update to the sensitivity affect-ing the response of the UV grism mainly below 2000 A 4The net continuum counts in the UV part of the individualspectra are very low so the first four and last four spectrawere summed which also improved the signal to noise (SN)of the weaker lines significantly No reddening correction hasbeen applied to the spectra

25 XMM-Newton X-ray and UV observations

V407 Lup was observed by XMM-Newton from 2017 March11 1145 to 1708 UT 1685 days post-eruption with anexposure duration of 23 ks (Ness et al 2017) The XMM-

Newton observatory consists of five different X-ray in-struments behind three mirrors plus an optical monitor(OMMason et al 2001 Talavera 2009) which all observesimultaneously For a full description of the X-ray instru-ments onboard of XMM-Newton see Jansen et al (2001)den Herder et al (2001) Struder et al (2001) Turner et al(2001) and Aschenbach (2002)

In this work we only used the spectra and light-curvesfrom the Reflection Grating Spectrometer (RGSden Herderet al 2001) the light-curves from the European PhotonImaging Camera (EPIC)pn (Struder et al 2001) and theOM light-curvesgrism spectra (see Table D5 for a log of theobservations)

The OM took five exposures one of which with the vis-ible grism provided a spectrum between 3000 and 6000 A(XMM-Newton could not take a U -grism spectrum becauseof a contaminating nearby star so we only obtained a V -grism spectrum) and four with Science User Defined imag-ing plus fast mode which provided a UV light-curve withthe uvw1 (λ sim 2910plusmn 830 A) filter

4 seehttpmssluclacuknpmkGrism

The RGS provides a spectral range of 6 ndash 38 A fully cov-ering the Wien tail of SSS spectra (30 ndash 80 eV) while theRayleigh Jeans tail is not visible (owing to interstellar ab-sorption) It also provides X-ray light-curves in the sameenergy range RGS consists of two instruments RGS1 andRGS2 One of the nine CCDs in RGS2 is suffering from sometechnical problems We Combine the RGS1 and RGS2 spec-tra in order to overcome the problem of the broken CCD inthe RGS2 instrument The EPICpn instrument was oper-ated in Timing Mode with medium filter providing anotherX-ray light-curve

26 Chandra X-ray observations

V407 Lup was observed with the Chandra X-ray Observa-

tory (hereafter Chandra Weisskopf et al 2000) using theHigh Resolution Camera (HRC Murray et al 1997) andthe Low Energy Transmission Grating (LETG Brinkmanet al 2000ab) on 2018 August 30 (day 3406) with an ex-posure time of 34 ks The LETG instrument provides a spec-trum over a range of sim 12 ndash 175 A (007 ndash 1033 keV) how-ever most of the flux from nova V407 Lup is in the range15 ndash 50 A (024 ndash 082 keV) The light-curve provided by theHRC instrument has a range of 006 ndash 10 keV

3 PHOTOMETRIC RESULTS AND ANALYSIS

31 Optical light-curve parameters

Several parameters characterize nova light-curves includingthe rise rate the rise time to maximum light the maximumlight the decline rate and the decline behaviour (see egHounsell et al 2010 Cao et al 2012) Nova V407 Lup wasnot extensively observed during its rise to maximum

Based on V -band and visual (Vis) measurements fromSMARTS and the American Association of Variable StarObservers (AAVSO)5 we see that the nova reached V =68 04 d after t0 Then it was reported at V = 64 on HJD245765624 (Prieto 2016) and Vis = 63 on HJD 245765657reaching Vis = 56 on HJD 245765690 (see Table C1) Halfa day later the nova was at V = 633 and Vis = 65 thendropped to V = 785 in around two days indicating thatthe decline had started Hence we assume HJD 245765690as tmax (day 14)

Caution is required in interpreting magnitudes of no-vae This is particularly so for Vis where Hα may con-tribute to the flux detected by eye but not to CCD V -magnitudes Furthermore normal transformation relationsfor CCD magnitudes cannot be used for objects with strongemission lines Unfortunately no spectra were taken aroundmaximum light so the contribution and development of lineemission is unknown until later times (day 5 Izzo et al2018) In the following we assume that maximum light wasVmax 6 60 on HJD 245765690 Since novae do not usu-ally have strong line emission at maximum light (van denBergh amp Younger 1987) and given that simultaneous V andVis measurements show very similar magnitudes before andafter maximum light it may be that V sim Vis =56 onHJD 245765690 is a reasonable estimate However we use

5 httpswwwaavsoorg

nova V407 Lup 5

Figure 1 V -band light-curve using data from SMARTS (green circles) AAVSO (red squares) and ATels from Stanek et al (2016)(yellow circle) Chen et al (2016) (blue stars) and Prieto (2016) (cyan diamond) A zoom-in plot around the first sim 25 days is added forclarity Within this plot the blue dashed line represents a power-law fit to the data (before day 25) excluding the discovery measurementthe black horizontal dashed lines represent from top to bottom Vmax Vmax + 2 and Vmax + 3 respectively the black vertical dashedlines represent from left to right tmax t2 and t3 respectively

Vmax 6 60 for the rest of the analysis to avoid overestimat-ing the maximum brightness

We construct a V -band light-curve by combining thepublished photometry from different telescopes and instru-ments and the SMARTS data (see Fig 1) noting the abovecaveat This shows a rapid rise and a sharp peak followed bya fast decline Such behaviour is characteristic of the S-classnova light-curves (Strope Schaefer amp Henden 2010) whichis considered stereotypical for a nova However the broad-band light-curves show day-to day variability after t3 similarto that of an O-class light-curves (see Fig B1) It is worthnoting that there is no presence for a plateau in the tail ofthe light-curve (after sim 3 ndash 6mag from maximum) similar tothat seen in recurrent novae (see Schaefer 2010)

Although there is a gap in the light-curve between day26 and day 107 (due to solar constraints) we fit a power-lawto the light-curve (only early decline - before day 25) andderive t2 6 29plusmn 05 d t3 6 68 plusmn 10 d and a power index ofsim 0166 We took into account the use of measurements fromdifferent instruments by increasing (times 2) the uncertainty ont2 and t3

With t2 6 29 d and a decline rate of sim 069mag dminus1

over t2 nova V407 Lup is a ldquovery-fastrdquo nova in the classifi-

6 Izzo et al (2018) have derived longer values of t2 and t3 This

was due to misinterpretation of their light-curve (private commu-nication with L Izzo)

Figure 2 The UVOT uvw2 (λcentral = 1928 A) light-curve plot-ted against day since t0

cation of Payne-Gaposchkin (1964) and is one of the fastestknown examples Only a few other novae have shown adecline time t2 30 days including M31N 2008-12a USco V1500 Cyg V838 Her V394 CrA and V4160 Sgr (seeegYoung et al 1976 Schaefer 2010 Munari et al 2011and table 5 in Darnley et al 2016)

6 Aydi et al

Figure 3 First panel the XMM-Newton EPICpn X-ray light-curve Second panel the RGS1 X-ray light-curve The X-ray light-curvesshow two broad (sim 2 ndash 3 ks wide) dips separated by 126 ks consistent with the 357 h modulation seen in the Swift-UVOT data (seeSection 34) Third panel the uvw1 OM fast mode (black) and imaging mode (red) light-curves The first OM exposure is a grismspectrum hence the light-curves only start at sim 5 ks The blue line represents a best fit sine curve to the OM data with a period of376plusmn03 h This period is consistent within uncertainty with the period seen in the Swift-UVOT data (see Section 34 for further

discussion on the timing analysis) The y-axis in all the panels are counts per second

32 The Swift UVOT light-curve

The UVOT uvw2 light-curve (Fig 2) declined slowly afterday 150 before flattening off at around a uvw2 magnitude of142 by around day 200 with signs of a 01mag variabilitysuperimposed on the overall decline We find periodicities inthe light-curve which we discuss in Section 34 The UV datarebrightened after day 300 to reach a uvw2 magnitude of137 around day 320 Then the brightness started to declineagain reaching a uvw2 magnitude of 149 at day sim 385Fig B2 shows a direct comparison between the Swift X-raySwift UV and SMARTS optical light-curves

33 The XMM-Newton and Chandra UV and

X-ray light-curves

The XMM-Newton RGS1 X-ray light-curve the EPICpnX-ray light-curve and the OM UV fast and imaging modeslight-curves are all presented in Fig 3 The XMM-Newton

RGS1 light-curve shows evidence of variability on differenttimescales This includes two broad (sim 2 ndash 3 ks wide) dipsseparated by sim 126 ks sim 35 h The OM fast mode UV light-curve shows a variation characterized by a 376plusmn03 h periodwhich is also consistent (within uncertainty) with the dura-tion between the dips in the RGS1 light-curve Howeverno other periodicity is found in the OM light-curve TheChandra HRC light-curve (Fig 4) also shows evidence forshort-term variability which we discuss in Section 34

Figure 4 The Chandra High Resolution Camera (HRC) X-ray

light-curve The HRC energy range is 006 ndash 10 keV There is aclear short-term variability in the light-curve and it is discussed

in Section 34

34 Timing analysis

Since the SwiftUVOT uvw2 light-curve (Fig 2) shows evi-dence for intrinsic variability superimposed on the long termdecline we searched for periodicities in the data Fig 5shows a Lomb-Scargle periodogram (LSP) of the barycen-tric corrected uvw2 light-curve after subtracting a third or-der polynomial to remove the long term trend The mostsignificant peaks occur at periods of 35731 plusmn 00014 h and

nova V407 Lup 7

Figure 5 Upper panel the Lomb Scargle periodogram of thedetrended uvw2 light-curve revealing significant modulation atfrequencies of 67168 dminus1 (777times10minus5 sminus1 P = 3573 h) and217696 dminus1 (252 times 10minus4 sminus1 P = 1102 h with estimated un-certainties of 00003 dminus1) marked with red dashed lines Lowerpanel periodogram of the observing window showing Swift rsquos or-

bital frequency (dotted line)

11024plusmn 00002 h which are aliases of each other caused bySwift rsquos 16 h orbit One of these periods might represent theorbital period of the binary (Porb) The low cadence of theUVOT data means we cannot break the degeneracy betweenthe periods (the high-cadence observation interval betweendays 307 and 311 failed to resolve this issue) The amplitudeof the modulation is approximately 01mag

We performed a similar Lomb-Scargle analysis of theXMM-Newton RGS1 and Chandra HRC data (see Fig 6)The light-curve of the former shows considerable flux vari-ations while the flux in the latter is approximately con-stant over the duration of the observation The XMM-

Newton periodogram is dominated by a low frequency ofΩ = 777times10minus5 sminus1 (357 h) This signal only covers sim twocycles of the 357 h period and cannot be believed in iso-lation However as this signal is consistent with the longperiod seen in the UVOT data we identify it as the same357 h modulation

The long-timescale modulation and variability seen inthe XMM-Newton light-curves the time between the dipsand the weak modulation observed in the OM light-curve(see Figs 3 and 6) are all consistent with the 357 h signalseen in the Swift UV light-curve In addition the absenceof any variability in the XMM-Newton RGS and OM light-curves related to the 11 h signal (seen in the Swift UV light-curve) has the potential to rule out the modulation at thisshort period Therefore we assume that the 357 h is mostlikely the Porb of the binary

Ignoring this long-term variability seen in the low fre-quency part of XMM-Newton periodogram we find signifi-cant signals at ω ≃ 177times10minus3 sminus1 in both (XMM-Newton

and Chandra) light-curves More precisely the periodicitieswe found are at 56504plusmn 068 s for the Chandra HRC light-curve and 56496plusmn 050 s for the XMM-Newton RGS light-curve The SwiftXRT observations are almost all too shortto usefully constrain the modulation at this short period

Figure 6 Lomb Scargle periodograms of the XMM-NewtonRGS1 (top) and the Chandra HRC (bottom) light-curves Vari-ous frequencies are indicated where Ω = 1Porb and ω = 1Pspin

(see text for more details) We present zoom-in plots around the177times10minus3 sminus1 in each periodogram

Figure 7 Top from top to bottom soft RGS1 light-curve hard

RGS1 light-curve and the hardness ratio The energy bandsfor the soft and hard light-curves are 235 ndash 370 A (03325 ndash

0528 keV) and 150 ndash 235 A (0528 ndash 0827 keV) respectively Bot-tom same as top but all folded at the 565 s spin period andnormalized to their respective mean rates Therefore the y-axisrepresents the relative amplitude of the modulation in both thesoft and hard bands

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 5: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 3

uniformly 50 s thereafter The procedure followed for reduc-ing the SMARTS photometry is detailed in Walter et al(2012) Fig B1 represents the SMARTS BVRI and JHK

observations See Table D1 for a log of the observations

22 SALT high-resolution Echelle spectroscopy

We used the SALT High Resolution Spectrograph (HRSBarnes et al 2008 Bramall et al 2010 Bramall et al 2012Crause et al 2014) to obtain observations on the nights of2017 March 6 April 20 May 31 June 13 29 July 05 24 29and August 06 (respectively days 164 209 250 263 279285 304 309 and 317) HRS a dual beam fibre-fed Echellespectrograph housed in a temperature stabilized vacuumtank was used in the low-resolution (LR) mode to obtainall the observations This mode provides two spectral rangesblue (3800 ndash 5550 A) and red (5450 ndash 9000 A) at a resolutionof R sim 15000 A weekly set of HRS calibrations includingfour ThAr + Ar arc spectra and four spectral flats is ob-tained in all the modes (low medium and high-resolution)All the HRS observations are 1800 s exposures See Table D2for a log of the observations

The primary reduction was conducted using the SALTscience pipeline (Crawford et al 2010) which includes over-scan correction bias subtraction and gain correction Therest of the reduction and relative flux calibration wasdone using the standard MIDAS FEROS (Stahl Kaufer ampTubbesing 1999) and echelle (Ballester 1992) packages Thereduction procedure is described in detail in Kniazev Gvara-madze amp Berdnikov (2016) Note that absolute flux calibra-tion is not feasible with SALT data As part of the SALTdesign the effective area of the telescope constantly changeswith the moving pupil during the track and exposures

23 SOAR medium-resolution spectroscopy

We performed optical spectroscopy of the nova on 2017 June20 (day 270) using the Goodman spectrograph (ClemensCrain amp Anderson 2004) on the 41m Southern Astrophys-ical Research (SOAR) telescope We obtained a single ex-posure of 600 s using a 600 lmmminus1 grating and a 107primeprime slitto provide a resolution of sim 13 A over the range of 3200 ndash6800 A The spectrum was reduced using the standard rou-tine and a relative flux calibration has been applied

We also obtained two spectra using the Goodman spec-trograph on the nights of 2018 January 20 and 21 (days484 and 485) each of 20min exposure For these two ob-servations we used a setup with a 2100 lmmminus1 grating anda 095primeprimeslit yielding a resolution of about 088 A full widthat half maximum (FWHM) (56 km sminus1) over a wavelengthrange of 4280 ndash 4950 AA The spectra were reduced and op-timally extracted in the usual manner

24 Swift X-ray and UV observations

241 Swift XRT observations

The Neil Gehrels Swift Observatory (hereafter Swift Gehrels et al 2004) first observed V407 Lup on 2016 Septem-ber 26 two days after the discovery The high optical bright-ness of the nova at this time meant that the UVOpticalTelescope (UVOT Roming et al 2005) could not be used

and the X-ray Telescope (XRT Burrows et al 2005) neededto be operated in Windowed Timing (WT) mode for theinitial observation in order to minimize the effects of op-tical loading whereby a large number of optical photonscan pile-up to appear as a false X-ray signature2 No X-raysource was detected at this time (lt 002 count sminus1 grade 0ndash single pixel events) or in the individual subsequent dailyobservations taken in Photon Counting (PC) mode between2016 October 2 and 9 (days 8 ndash 15) Coadding these PC datathere is a source detected at the 99 per cent confidence levelwith a count rate of (14+05

minus04) times10minus3 count sminus1 (grade 0)There was also a suggestion of a detection in the October10 (day 16) data alone These detections could not be con-firmed at a higher significance before V407 Lup became tooclose to the Sun for Swift to observe on 2016 October 13(day 19)

Observations recommenced on 2017 February 21 (day150) finding a bright supersoft X-ray source (Beardmoreet al 2017) with a count rate of 561plusmn 03 count sminus1 Bythis time the UVOT could also be safely operated and a UVsource with uvw2 = 1349plusmn 002mag was measured Dailyobservations were performed between 2017 February 22 and28 (days 151 ndash 157) followed by observations approximatelyevery two to three days until 2017 April 24 (day 212) Obser-vations were continued though with a slightly lower cadenceof every four days until 2017 September 10 (day 351) withthe cadence then decreasing to every eight days until 2017October 13 (day 384) when the Sun constraint began againA final dataset was obtained once the source again emergedfrom the Sun constraint on 2018 January 23 (day 486) Allobservations were typically 05-1 ks in duration with the ex-ception of the final observation which was 3 ks In addition tothis regular monitoring a high cadence campaign was per-formed between 2017 July 4 ndash 8 (days 283 ndash 287) aimed atpinning down the periodicity seen in the UVOT light-curve(see Section 34) In this case sim 500 s snapshots were takenapproximately every 6 h Because of complications causedby a nearby bright star no UVOT data were actually col-lected and so the campaign was repeated from 2017 July28 ndashAugust 1 (days 307 ndash 311) using an offset pointing toavoid this problem The Swift XRT observations log is givenin Table D3

The Swift data were processed with the standardHEASoft tools (version 620)3 and analysed using themost up-to-date calibration files All observations be-tween 2017 February 21 and August 11 (days 150 and321 post-eruption) inclusive were obtained using WTmode because of the high brightness of the X-raysource These data were extracted using a circular re-gion of a radius of 20 pixels (1 pixel = 236 arcsec)for the source and a background annulus as describedat httpwwwswiftacukanalysisxrtbackscalphp Ob-servations from 2017 August 15 (day 325) onwards weretaken with the XRT in PC mode the first three of thesesuffered from pile-up so an annulus (outer radius 30 pix-els inner exclusion radius decreasing from seven to three

2 httpwwwswiftacukanalysisxrtoptical_loading

php3 httpsheasarcgsfcnasagovFTPsoftwareftools

releasearchiveRelease_Notes_620

4 Aydi et al

pixels) was used when extracting the source counts Thelater data were analysed using circular regions decreasingin radius from 20 to 10 pixels as the source further fadedBackground counts were estimated from near-by source-freecircular regions of 60 pixels radius

The X-ray spectra were binned to provide a minimum of1 count binminus1 to facilitate Cash statistic (Cash 1979) fittingwithin xspec (Arnaud 1996)

242 Swift UVOT observations

Observations with the Swift UVOT instrument startedonce the nova could be observed again after its passage be-hind the Sun On 2017 February 26 (day 156) the nova wasobserved in the uvw2 (λcentral = 1928 A) filter the next dayin the uvm2 (λcentral = 2246 A) filter Photometric observa-tions continued until 2017 October 13 (day 385) A log ofthe observations is given in Table D3

Weekly observations with the UV grism started on 2017March 7 until April 23 (day 165 until day 212) when theywere discontinued as the brightness was too low The UVgrism provides a spectral range of sim 1700 ndash 5100 A Thelog of the observations are given in Table D4 The eightgrism observations were reduced using the UVOTPY soft-ware (Kuin 2014) using the calibration described in Kuinet al (2015) with a recent update to the sensitivity affect-ing the response of the UV grism mainly below 2000 A 4The net continuum counts in the UV part of the individualspectra are very low so the first four and last four spectrawere summed which also improved the signal to noise (SN)of the weaker lines significantly No reddening correction hasbeen applied to the spectra

25 XMM-Newton X-ray and UV observations

V407 Lup was observed by XMM-Newton from 2017 March11 1145 to 1708 UT 1685 days post-eruption with anexposure duration of 23 ks (Ness et al 2017) The XMM-

Newton observatory consists of five different X-ray in-struments behind three mirrors plus an optical monitor(OMMason et al 2001 Talavera 2009) which all observesimultaneously For a full description of the X-ray instru-ments onboard of XMM-Newton see Jansen et al (2001)den Herder et al (2001) Struder et al (2001) Turner et al(2001) and Aschenbach (2002)

In this work we only used the spectra and light-curvesfrom the Reflection Grating Spectrometer (RGSden Herderet al 2001) the light-curves from the European PhotonImaging Camera (EPIC)pn (Struder et al 2001) and theOM light-curvesgrism spectra (see Table D5 for a log of theobservations)

The OM took five exposures one of which with the vis-ible grism provided a spectrum between 3000 and 6000 A(XMM-Newton could not take a U -grism spectrum becauseof a contaminating nearby star so we only obtained a V -grism spectrum) and four with Science User Defined imag-ing plus fast mode which provided a UV light-curve withthe uvw1 (λ sim 2910plusmn 830 A) filter

4 seehttpmssluclacuknpmkGrism

The RGS provides a spectral range of 6 ndash 38 A fully cov-ering the Wien tail of SSS spectra (30 ndash 80 eV) while theRayleigh Jeans tail is not visible (owing to interstellar ab-sorption) It also provides X-ray light-curves in the sameenergy range RGS consists of two instruments RGS1 andRGS2 One of the nine CCDs in RGS2 is suffering from sometechnical problems We Combine the RGS1 and RGS2 spec-tra in order to overcome the problem of the broken CCD inthe RGS2 instrument The EPICpn instrument was oper-ated in Timing Mode with medium filter providing anotherX-ray light-curve

26 Chandra X-ray observations

V407 Lup was observed with the Chandra X-ray Observa-

tory (hereafter Chandra Weisskopf et al 2000) using theHigh Resolution Camera (HRC Murray et al 1997) andthe Low Energy Transmission Grating (LETG Brinkmanet al 2000ab) on 2018 August 30 (day 3406) with an ex-posure time of 34 ks The LETG instrument provides a spec-trum over a range of sim 12 ndash 175 A (007 ndash 1033 keV) how-ever most of the flux from nova V407 Lup is in the range15 ndash 50 A (024 ndash 082 keV) The light-curve provided by theHRC instrument has a range of 006 ndash 10 keV

3 PHOTOMETRIC RESULTS AND ANALYSIS

31 Optical light-curve parameters

Several parameters characterize nova light-curves includingthe rise rate the rise time to maximum light the maximumlight the decline rate and the decline behaviour (see egHounsell et al 2010 Cao et al 2012) Nova V407 Lup wasnot extensively observed during its rise to maximum

Based on V -band and visual (Vis) measurements fromSMARTS and the American Association of Variable StarObservers (AAVSO)5 we see that the nova reached V =68 04 d after t0 Then it was reported at V = 64 on HJD245765624 (Prieto 2016) and Vis = 63 on HJD 245765657reaching Vis = 56 on HJD 245765690 (see Table C1) Halfa day later the nova was at V = 633 and Vis = 65 thendropped to V = 785 in around two days indicating thatthe decline had started Hence we assume HJD 245765690as tmax (day 14)

Caution is required in interpreting magnitudes of no-vae This is particularly so for Vis where Hα may con-tribute to the flux detected by eye but not to CCD V -magnitudes Furthermore normal transformation relationsfor CCD magnitudes cannot be used for objects with strongemission lines Unfortunately no spectra were taken aroundmaximum light so the contribution and development of lineemission is unknown until later times (day 5 Izzo et al2018) In the following we assume that maximum light wasVmax 6 60 on HJD 245765690 Since novae do not usu-ally have strong line emission at maximum light (van denBergh amp Younger 1987) and given that simultaneous V andVis measurements show very similar magnitudes before andafter maximum light it may be that V sim Vis =56 onHJD 245765690 is a reasonable estimate However we use

5 httpswwwaavsoorg

nova V407 Lup 5

Figure 1 V -band light-curve using data from SMARTS (green circles) AAVSO (red squares) and ATels from Stanek et al (2016)(yellow circle) Chen et al (2016) (blue stars) and Prieto (2016) (cyan diamond) A zoom-in plot around the first sim 25 days is added forclarity Within this plot the blue dashed line represents a power-law fit to the data (before day 25) excluding the discovery measurementthe black horizontal dashed lines represent from top to bottom Vmax Vmax + 2 and Vmax + 3 respectively the black vertical dashedlines represent from left to right tmax t2 and t3 respectively

Vmax 6 60 for the rest of the analysis to avoid overestimat-ing the maximum brightness

We construct a V -band light-curve by combining thepublished photometry from different telescopes and instru-ments and the SMARTS data (see Fig 1) noting the abovecaveat This shows a rapid rise and a sharp peak followed bya fast decline Such behaviour is characteristic of the S-classnova light-curves (Strope Schaefer amp Henden 2010) whichis considered stereotypical for a nova However the broad-band light-curves show day-to day variability after t3 similarto that of an O-class light-curves (see Fig B1) It is worthnoting that there is no presence for a plateau in the tail ofthe light-curve (after sim 3 ndash 6mag from maximum) similar tothat seen in recurrent novae (see Schaefer 2010)

Although there is a gap in the light-curve between day26 and day 107 (due to solar constraints) we fit a power-lawto the light-curve (only early decline - before day 25) andderive t2 6 29plusmn 05 d t3 6 68 plusmn 10 d and a power index ofsim 0166 We took into account the use of measurements fromdifferent instruments by increasing (times 2) the uncertainty ont2 and t3

With t2 6 29 d and a decline rate of sim 069mag dminus1

over t2 nova V407 Lup is a ldquovery-fastrdquo nova in the classifi-

6 Izzo et al (2018) have derived longer values of t2 and t3 This

was due to misinterpretation of their light-curve (private commu-nication with L Izzo)

Figure 2 The UVOT uvw2 (λcentral = 1928 A) light-curve plot-ted against day since t0

cation of Payne-Gaposchkin (1964) and is one of the fastestknown examples Only a few other novae have shown adecline time t2 30 days including M31N 2008-12a USco V1500 Cyg V838 Her V394 CrA and V4160 Sgr (seeegYoung et al 1976 Schaefer 2010 Munari et al 2011and table 5 in Darnley et al 2016)

6 Aydi et al

Figure 3 First panel the XMM-Newton EPICpn X-ray light-curve Second panel the RGS1 X-ray light-curve The X-ray light-curvesshow two broad (sim 2 ndash 3 ks wide) dips separated by 126 ks consistent with the 357 h modulation seen in the Swift-UVOT data (seeSection 34) Third panel the uvw1 OM fast mode (black) and imaging mode (red) light-curves The first OM exposure is a grismspectrum hence the light-curves only start at sim 5 ks The blue line represents a best fit sine curve to the OM data with a period of376plusmn03 h This period is consistent within uncertainty with the period seen in the Swift-UVOT data (see Section 34 for further

discussion on the timing analysis) The y-axis in all the panels are counts per second

32 The Swift UVOT light-curve

The UVOT uvw2 light-curve (Fig 2) declined slowly afterday 150 before flattening off at around a uvw2 magnitude of142 by around day 200 with signs of a 01mag variabilitysuperimposed on the overall decline We find periodicities inthe light-curve which we discuss in Section 34 The UV datarebrightened after day 300 to reach a uvw2 magnitude of137 around day 320 Then the brightness started to declineagain reaching a uvw2 magnitude of 149 at day sim 385Fig B2 shows a direct comparison between the Swift X-raySwift UV and SMARTS optical light-curves

33 The XMM-Newton and Chandra UV and

X-ray light-curves

The XMM-Newton RGS1 X-ray light-curve the EPICpnX-ray light-curve and the OM UV fast and imaging modeslight-curves are all presented in Fig 3 The XMM-Newton

RGS1 light-curve shows evidence of variability on differenttimescales This includes two broad (sim 2 ndash 3 ks wide) dipsseparated by sim 126 ks sim 35 h The OM fast mode UV light-curve shows a variation characterized by a 376plusmn03 h periodwhich is also consistent (within uncertainty) with the dura-tion between the dips in the RGS1 light-curve Howeverno other periodicity is found in the OM light-curve TheChandra HRC light-curve (Fig 4) also shows evidence forshort-term variability which we discuss in Section 34

Figure 4 The Chandra High Resolution Camera (HRC) X-ray

light-curve The HRC energy range is 006 ndash 10 keV There is aclear short-term variability in the light-curve and it is discussed

in Section 34

34 Timing analysis

Since the SwiftUVOT uvw2 light-curve (Fig 2) shows evi-dence for intrinsic variability superimposed on the long termdecline we searched for periodicities in the data Fig 5shows a Lomb-Scargle periodogram (LSP) of the barycen-tric corrected uvw2 light-curve after subtracting a third or-der polynomial to remove the long term trend The mostsignificant peaks occur at periods of 35731 plusmn 00014 h and

nova V407 Lup 7

Figure 5 Upper panel the Lomb Scargle periodogram of thedetrended uvw2 light-curve revealing significant modulation atfrequencies of 67168 dminus1 (777times10minus5 sminus1 P = 3573 h) and217696 dminus1 (252 times 10minus4 sminus1 P = 1102 h with estimated un-certainties of 00003 dminus1) marked with red dashed lines Lowerpanel periodogram of the observing window showing Swift rsquos or-

bital frequency (dotted line)

11024plusmn 00002 h which are aliases of each other caused bySwift rsquos 16 h orbit One of these periods might represent theorbital period of the binary (Porb) The low cadence of theUVOT data means we cannot break the degeneracy betweenthe periods (the high-cadence observation interval betweendays 307 and 311 failed to resolve this issue) The amplitudeof the modulation is approximately 01mag

We performed a similar Lomb-Scargle analysis of theXMM-Newton RGS1 and Chandra HRC data (see Fig 6)The light-curve of the former shows considerable flux vari-ations while the flux in the latter is approximately con-stant over the duration of the observation The XMM-

Newton periodogram is dominated by a low frequency ofΩ = 777times10minus5 sminus1 (357 h) This signal only covers sim twocycles of the 357 h period and cannot be believed in iso-lation However as this signal is consistent with the longperiod seen in the UVOT data we identify it as the same357 h modulation

The long-timescale modulation and variability seen inthe XMM-Newton light-curves the time between the dipsand the weak modulation observed in the OM light-curve(see Figs 3 and 6) are all consistent with the 357 h signalseen in the Swift UV light-curve In addition the absenceof any variability in the XMM-Newton RGS and OM light-curves related to the 11 h signal (seen in the Swift UV light-curve) has the potential to rule out the modulation at thisshort period Therefore we assume that the 357 h is mostlikely the Porb of the binary

Ignoring this long-term variability seen in the low fre-quency part of XMM-Newton periodogram we find signifi-cant signals at ω ≃ 177times10minus3 sminus1 in both (XMM-Newton

and Chandra) light-curves More precisely the periodicitieswe found are at 56504plusmn 068 s for the Chandra HRC light-curve and 56496plusmn 050 s for the XMM-Newton RGS light-curve The SwiftXRT observations are almost all too shortto usefully constrain the modulation at this short period

Figure 6 Lomb Scargle periodograms of the XMM-NewtonRGS1 (top) and the Chandra HRC (bottom) light-curves Vari-ous frequencies are indicated where Ω = 1Porb and ω = 1Pspin

(see text for more details) We present zoom-in plots around the177times10minus3 sminus1 in each periodogram

Figure 7 Top from top to bottom soft RGS1 light-curve hard

RGS1 light-curve and the hardness ratio The energy bandsfor the soft and hard light-curves are 235 ndash 370 A (03325 ndash

0528 keV) and 150 ndash 235 A (0528 ndash 0827 keV) respectively Bot-tom same as top but all folded at the 565 s spin period andnormalized to their respective mean rates Therefore the y-axisrepresents the relative amplitude of the modulation in both thesoft and hard bands

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 6: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

4 Aydi et al

pixels) was used when extracting the source counts Thelater data were analysed using circular regions decreasingin radius from 20 to 10 pixels as the source further fadedBackground counts were estimated from near-by source-freecircular regions of 60 pixels radius

The X-ray spectra were binned to provide a minimum of1 count binminus1 to facilitate Cash statistic (Cash 1979) fittingwithin xspec (Arnaud 1996)

242 Swift UVOT observations

Observations with the Swift UVOT instrument startedonce the nova could be observed again after its passage be-hind the Sun On 2017 February 26 (day 156) the nova wasobserved in the uvw2 (λcentral = 1928 A) filter the next dayin the uvm2 (λcentral = 2246 A) filter Photometric observa-tions continued until 2017 October 13 (day 385) A log ofthe observations is given in Table D3

Weekly observations with the UV grism started on 2017March 7 until April 23 (day 165 until day 212) when theywere discontinued as the brightness was too low The UVgrism provides a spectral range of sim 1700 ndash 5100 A Thelog of the observations are given in Table D4 The eightgrism observations were reduced using the UVOTPY soft-ware (Kuin 2014) using the calibration described in Kuinet al (2015) with a recent update to the sensitivity affect-ing the response of the UV grism mainly below 2000 A 4The net continuum counts in the UV part of the individualspectra are very low so the first four and last four spectrawere summed which also improved the signal to noise (SN)of the weaker lines significantly No reddening correction hasbeen applied to the spectra

25 XMM-Newton X-ray and UV observations

V407 Lup was observed by XMM-Newton from 2017 March11 1145 to 1708 UT 1685 days post-eruption with anexposure duration of 23 ks (Ness et al 2017) The XMM-

Newton observatory consists of five different X-ray in-struments behind three mirrors plus an optical monitor(OMMason et al 2001 Talavera 2009) which all observesimultaneously For a full description of the X-ray instru-ments onboard of XMM-Newton see Jansen et al (2001)den Herder et al (2001) Struder et al (2001) Turner et al(2001) and Aschenbach (2002)

In this work we only used the spectra and light-curvesfrom the Reflection Grating Spectrometer (RGSden Herderet al 2001) the light-curves from the European PhotonImaging Camera (EPIC)pn (Struder et al 2001) and theOM light-curvesgrism spectra (see Table D5 for a log of theobservations)

The OM took five exposures one of which with the vis-ible grism provided a spectrum between 3000 and 6000 A(XMM-Newton could not take a U -grism spectrum becauseof a contaminating nearby star so we only obtained a V -grism spectrum) and four with Science User Defined imag-ing plus fast mode which provided a UV light-curve withthe uvw1 (λ sim 2910plusmn 830 A) filter

4 seehttpmssluclacuknpmkGrism

The RGS provides a spectral range of 6 ndash 38 A fully cov-ering the Wien tail of SSS spectra (30 ndash 80 eV) while theRayleigh Jeans tail is not visible (owing to interstellar ab-sorption) It also provides X-ray light-curves in the sameenergy range RGS consists of two instruments RGS1 andRGS2 One of the nine CCDs in RGS2 is suffering from sometechnical problems We Combine the RGS1 and RGS2 spec-tra in order to overcome the problem of the broken CCD inthe RGS2 instrument The EPICpn instrument was oper-ated in Timing Mode with medium filter providing anotherX-ray light-curve

26 Chandra X-ray observations

V407 Lup was observed with the Chandra X-ray Observa-

tory (hereafter Chandra Weisskopf et al 2000) using theHigh Resolution Camera (HRC Murray et al 1997) andthe Low Energy Transmission Grating (LETG Brinkmanet al 2000ab) on 2018 August 30 (day 3406) with an ex-posure time of 34 ks The LETG instrument provides a spec-trum over a range of sim 12 ndash 175 A (007 ndash 1033 keV) how-ever most of the flux from nova V407 Lup is in the range15 ndash 50 A (024 ndash 082 keV) The light-curve provided by theHRC instrument has a range of 006 ndash 10 keV

3 PHOTOMETRIC RESULTS AND ANALYSIS

31 Optical light-curve parameters

Several parameters characterize nova light-curves includingthe rise rate the rise time to maximum light the maximumlight the decline rate and the decline behaviour (see egHounsell et al 2010 Cao et al 2012) Nova V407 Lup wasnot extensively observed during its rise to maximum

Based on V -band and visual (Vis) measurements fromSMARTS and the American Association of Variable StarObservers (AAVSO)5 we see that the nova reached V =68 04 d after t0 Then it was reported at V = 64 on HJD245765624 (Prieto 2016) and Vis = 63 on HJD 245765657reaching Vis = 56 on HJD 245765690 (see Table C1) Halfa day later the nova was at V = 633 and Vis = 65 thendropped to V = 785 in around two days indicating thatthe decline had started Hence we assume HJD 245765690as tmax (day 14)

Caution is required in interpreting magnitudes of no-vae This is particularly so for Vis where Hα may con-tribute to the flux detected by eye but not to CCD V -magnitudes Furthermore normal transformation relationsfor CCD magnitudes cannot be used for objects with strongemission lines Unfortunately no spectra were taken aroundmaximum light so the contribution and development of lineemission is unknown until later times (day 5 Izzo et al2018) In the following we assume that maximum light wasVmax 6 60 on HJD 245765690 Since novae do not usu-ally have strong line emission at maximum light (van denBergh amp Younger 1987) and given that simultaneous V andVis measurements show very similar magnitudes before andafter maximum light it may be that V sim Vis =56 onHJD 245765690 is a reasonable estimate However we use

5 httpswwwaavsoorg

nova V407 Lup 5

Figure 1 V -band light-curve using data from SMARTS (green circles) AAVSO (red squares) and ATels from Stanek et al (2016)(yellow circle) Chen et al (2016) (blue stars) and Prieto (2016) (cyan diamond) A zoom-in plot around the first sim 25 days is added forclarity Within this plot the blue dashed line represents a power-law fit to the data (before day 25) excluding the discovery measurementthe black horizontal dashed lines represent from top to bottom Vmax Vmax + 2 and Vmax + 3 respectively the black vertical dashedlines represent from left to right tmax t2 and t3 respectively

Vmax 6 60 for the rest of the analysis to avoid overestimat-ing the maximum brightness

We construct a V -band light-curve by combining thepublished photometry from different telescopes and instru-ments and the SMARTS data (see Fig 1) noting the abovecaveat This shows a rapid rise and a sharp peak followed bya fast decline Such behaviour is characteristic of the S-classnova light-curves (Strope Schaefer amp Henden 2010) whichis considered stereotypical for a nova However the broad-band light-curves show day-to day variability after t3 similarto that of an O-class light-curves (see Fig B1) It is worthnoting that there is no presence for a plateau in the tail ofthe light-curve (after sim 3 ndash 6mag from maximum) similar tothat seen in recurrent novae (see Schaefer 2010)

Although there is a gap in the light-curve between day26 and day 107 (due to solar constraints) we fit a power-lawto the light-curve (only early decline - before day 25) andderive t2 6 29plusmn 05 d t3 6 68 plusmn 10 d and a power index ofsim 0166 We took into account the use of measurements fromdifferent instruments by increasing (times 2) the uncertainty ont2 and t3

With t2 6 29 d and a decline rate of sim 069mag dminus1

over t2 nova V407 Lup is a ldquovery-fastrdquo nova in the classifi-

6 Izzo et al (2018) have derived longer values of t2 and t3 This

was due to misinterpretation of their light-curve (private commu-nication with L Izzo)

Figure 2 The UVOT uvw2 (λcentral = 1928 A) light-curve plot-ted against day since t0

cation of Payne-Gaposchkin (1964) and is one of the fastestknown examples Only a few other novae have shown adecline time t2 30 days including M31N 2008-12a USco V1500 Cyg V838 Her V394 CrA and V4160 Sgr (seeegYoung et al 1976 Schaefer 2010 Munari et al 2011and table 5 in Darnley et al 2016)

6 Aydi et al

Figure 3 First panel the XMM-Newton EPICpn X-ray light-curve Second panel the RGS1 X-ray light-curve The X-ray light-curvesshow two broad (sim 2 ndash 3 ks wide) dips separated by 126 ks consistent with the 357 h modulation seen in the Swift-UVOT data (seeSection 34) Third panel the uvw1 OM fast mode (black) and imaging mode (red) light-curves The first OM exposure is a grismspectrum hence the light-curves only start at sim 5 ks The blue line represents a best fit sine curve to the OM data with a period of376plusmn03 h This period is consistent within uncertainty with the period seen in the Swift-UVOT data (see Section 34 for further

discussion on the timing analysis) The y-axis in all the panels are counts per second

32 The Swift UVOT light-curve

The UVOT uvw2 light-curve (Fig 2) declined slowly afterday 150 before flattening off at around a uvw2 magnitude of142 by around day 200 with signs of a 01mag variabilitysuperimposed on the overall decline We find periodicities inthe light-curve which we discuss in Section 34 The UV datarebrightened after day 300 to reach a uvw2 magnitude of137 around day 320 Then the brightness started to declineagain reaching a uvw2 magnitude of 149 at day sim 385Fig B2 shows a direct comparison between the Swift X-raySwift UV and SMARTS optical light-curves

33 The XMM-Newton and Chandra UV and

X-ray light-curves

The XMM-Newton RGS1 X-ray light-curve the EPICpnX-ray light-curve and the OM UV fast and imaging modeslight-curves are all presented in Fig 3 The XMM-Newton

RGS1 light-curve shows evidence of variability on differenttimescales This includes two broad (sim 2 ndash 3 ks wide) dipsseparated by sim 126 ks sim 35 h The OM fast mode UV light-curve shows a variation characterized by a 376plusmn03 h periodwhich is also consistent (within uncertainty) with the dura-tion between the dips in the RGS1 light-curve Howeverno other periodicity is found in the OM light-curve TheChandra HRC light-curve (Fig 4) also shows evidence forshort-term variability which we discuss in Section 34

Figure 4 The Chandra High Resolution Camera (HRC) X-ray

light-curve The HRC energy range is 006 ndash 10 keV There is aclear short-term variability in the light-curve and it is discussed

in Section 34

34 Timing analysis

Since the SwiftUVOT uvw2 light-curve (Fig 2) shows evi-dence for intrinsic variability superimposed on the long termdecline we searched for periodicities in the data Fig 5shows a Lomb-Scargle periodogram (LSP) of the barycen-tric corrected uvw2 light-curve after subtracting a third or-der polynomial to remove the long term trend The mostsignificant peaks occur at periods of 35731 plusmn 00014 h and

nova V407 Lup 7

Figure 5 Upper panel the Lomb Scargle periodogram of thedetrended uvw2 light-curve revealing significant modulation atfrequencies of 67168 dminus1 (777times10minus5 sminus1 P = 3573 h) and217696 dminus1 (252 times 10minus4 sminus1 P = 1102 h with estimated un-certainties of 00003 dminus1) marked with red dashed lines Lowerpanel periodogram of the observing window showing Swift rsquos or-

bital frequency (dotted line)

11024plusmn 00002 h which are aliases of each other caused bySwift rsquos 16 h orbit One of these periods might represent theorbital period of the binary (Porb) The low cadence of theUVOT data means we cannot break the degeneracy betweenthe periods (the high-cadence observation interval betweendays 307 and 311 failed to resolve this issue) The amplitudeof the modulation is approximately 01mag

We performed a similar Lomb-Scargle analysis of theXMM-Newton RGS1 and Chandra HRC data (see Fig 6)The light-curve of the former shows considerable flux vari-ations while the flux in the latter is approximately con-stant over the duration of the observation The XMM-

Newton periodogram is dominated by a low frequency ofΩ = 777times10minus5 sminus1 (357 h) This signal only covers sim twocycles of the 357 h period and cannot be believed in iso-lation However as this signal is consistent with the longperiod seen in the UVOT data we identify it as the same357 h modulation

The long-timescale modulation and variability seen inthe XMM-Newton light-curves the time between the dipsand the weak modulation observed in the OM light-curve(see Figs 3 and 6) are all consistent with the 357 h signalseen in the Swift UV light-curve In addition the absenceof any variability in the XMM-Newton RGS and OM light-curves related to the 11 h signal (seen in the Swift UV light-curve) has the potential to rule out the modulation at thisshort period Therefore we assume that the 357 h is mostlikely the Porb of the binary

Ignoring this long-term variability seen in the low fre-quency part of XMM-Newton periodogram we find signifi-cant signals at ω ≃ 177times10minus3 sminus1 in both (XMM-Newton

and Chandra) light-curves More precisely the periodicitieswe found are at 56504plusmn 068 s for the Chandra HRC light-curve and 56496plusmn 050 s for the XMM-Newton RGS light-curve The SwiftXRT observations are almost all too shortto usefully constrain the modulation at this short period

Figure 6 Lomb Scargle periodograms of the XMM-NewtonRGS1 (top) and the Chandra HRC (bottom) light-curves Vari-ous frequencies are indicated where Ω = 1Porb and ω = 1Pspin

(see text for more details) We present zoom-in plots around the177times10minus3 sminus1 in each periodogram

Figure 7 Top from top to bottom soft RGS1 light-curve hard

RGS1 light-curve and the hardness ratio The energy bandsfor the soft and hard light-curves are 235 ndash 370 A (03325 ndash

0528 keV) and 150 ndash 235 A (0528 ndash 0827 keV) respectively Bot-tom same as top but all folded at the 565 s spin period andnormalized to their respective mean rates Therefore the y-axisrepresents the relative amplitude of the modulation in both thesoft and hard bands

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 7: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 5

Figure 1 V -band light-curve using data from SMARTS (green circles) AAVSO (red squares) and ATels from Stanek et al (2016)(yellow circle) Chen et al (2016) (blue stars) and Prieto (2016) (cyan diamond) A zoom-in plot around the first sim 25 days is added forclarity Within this plot the blue dashed line represents a power-law fit to the data (before day 25) excluding the discovery measurementthe black horizontal dashed lines represent from top to bottom Vmax Vmax + 2 and Vmax + 3 respectively the black vertical dashedlines represent from left to right tmax t2 and t3 respectively

Vmax 6 60 for the rest of the analysis to avoid overestimat-ing the maximum brightness

We construct a V -band light-curve by combining thepublished photometry from different telescopes and instru-ments and the SMARTS data (see Fig 1) noting the abovecaveat This shows a rapid rise and a sharp peak followed bya fast decline Such behaviour is characteristic of the S-classnova light-curves (Strope Schaefer amp Henden 2010) whichis considered stereotypical for a nova However the broad-band light-curves show day-to day variability after t3 similarto that of an O-class light-curves (see Fig B1) It is worthnoting that there is no presence for a plateau in the tail ofthe light-curve (after sim 3 ndash 6mag from maximum) similar tothat seen in recurrent novae (see Schaefer 2010)

Although there is a gap in the light-curve between day26 and day 107 (due to solar constraints) we fit a power-lawto the light-curve (only early decline - before day 25) andderive t2 6 29plusmn 05 d t3 6 68 plusmn 10 d and a power index ofsim 0166 We took into account the use of measurements fromdifferent instruments by increasing (times 2) the uncertainty ont2 and t3

With t2 6 29 d and a decline rate of sim 069mag dminus1

over t2 nova V407 Lup is a ldquovery-fastrdquo nova in the classifi-

6 Izzo et al (2018) have derived longer values of t2 and t3 This

was due to misinterpretation of their light-curve (private commu-nication with L Izzo)

Figure 2 The UVOT uvw2 (λcentral = 1928 A) light-curve plot-ted against day since t0

cation of Payne-Gaposchkin (1964) and is one of the fastestknown examples Only a few other novae have shown adecline time t2 30 days including M31N 2008-12a USco V1500 Cyg V838 Her V394 CrA and V4160 Sgr (seeegYoung et al 1976 Schaefer 2010 Munari et al 2011and table 5 in Darnley et al 2016)

6 Aydi et al

Figure 3 First panel the XMM-Newton EPICpn X-ray light-curve Second panel the RGS1 X-ray light-curve The X-ray light-curvesshow two broad (sim 2 ndash 3 ks wide) dips separated by 126 ks consistent with the 357 h modulation seen in the Swift-UVOT data (seeSection 34) Third panel the uvw1 OM fast mode (black) and imaging mode (red) light-curves The first OM exposure is a grismspectrum hence the light-curves only start at sim 5 ks The blue line represents a best fit sine curve to the OM data with a period of376plusmn03 h This period is consistent within uncertainty with the period seen in the Swift-UVOT data (see Section 34 for further

discussion on the timing analysis) The y-axis in all the panels are counts per second

32 The Swift UVOT light-curve

The UVOT uvw2 light-curve (Fig 2) declined slowly afterday 150 before flattening off at around a uvw2 magnitude of142 by around day 200 with signs of a 01mag variabilitysuperimposed on the overall decline We find periodicities inthe light-curve which we discuss in Section 34 The UV datarebrightened after day 300 to reach a uvw2 magnitude of137 around day 320 Then the brightness started to declineagain reaching a uvw2 magnitude of 149 at day sim 385Fig B2 shows a direct comparison between the Swift X-raySwift UV and SMARTS optical light-curves

33 The XMM-Newton and Chandra UV and

X-ray light-curves

The XMM-Newton RGS1 X-ray light-curve the EPICpnX-ray light-curve and the OM UV fast and imaging modeslight-curves are all presented in Fig 3 The XMM-Newton

RGS1 light-curve shows evidence of variability on differenttimescales This includes two broad (sim 2 ndash 3 ks wide) dipsseparated by sim 126 ks sim 35 h The OM fast mode UV light-curve shows a variation characterized by a 376plusmn03 h periodwhich is also consistent (within uncertainty) with the dura-tion between the dips in the RGS1 light-curve Howeverno other periodicity is found in the OM light-curve TheChandra HRC light-curve (Fig 4) also shows evidence forshort-term variability which we discuss in Section 34

Figure 4 The Chandra High Resolution Camera (HRC) X-ray

light-curve The HRC energy range is 006 ndash 10 keV There is aclear short-term variability in the light-curve and it is discussed

in Section 34

34 Timing analysis

Since the SwiftUVOT uvw2 light-curve (Fig 2) shows evi-dence for intrinsic variability superimposed on the long termdecline we searched for periodicities in the data Fig 5shows a Lomb-Scargle periodogram (LSP) of the barycen-tric corrected uvw2 light-curve after subtracting a third or-der polynomial to remove the long term trend The mostsignificant peaks occur at periods of 35731 plusmn 00014 h and

nova V407 Lup 7

Figure 5 Upper panel the Lomb Scargle periodogram of thedetrended uvw2 light-curve revealing significant modulation atfrequencies of 67168 dminus1 (777times10minus5 sminus1 P = 3573 h) and217696 dminus1 (252 times 10minus4 sminus1 P = 1102 h with estimated un-certainties of 00003 dminus1) marked with red dashed lines Lowerpanel periodogram of the observing window showing Swift rsquos or-

bital frequency (dotted line)

11024plusmn 00002 h which are aliases of each other caused bySwift rsquos 16 h orbit One of these periods might represent theorbital period of the binary (Porb) The low cadence of theUVOT data means we cannot break the degeneracy betweenthe periods (the high-cadence observation interval betweendays 307 and 311 failed to resolve this issue) The amplitudeof the modulation is approximately 01mag

We performed a similar Lomb-Scargle analysis of theXMM-Newton RGS1 and Chandra HRC data (see Fig 6)The light-curve of the former shows considerable flux vari-ations while the flux in the latter is approximately con-stant over the duration of the observation The XMM-

Newton periodogram is dominated by a low frequency ofΩ = 777times10minus5 sminus1 (357 h) This signal only covers sim twocycles of the 357 h period and cannot be believed in iso-lation However as this signal is consistent with the longperiod seen in the UVOT data we identify it as the same357 h modulation

The long-timescale modulation and variability seen inthe XMM-Newton light-curves the time between the dipsand the weak modulation observed in the OM light-curve(see Figs 3 and 6) are all consistent with the 357 h signalseen in the Swift UV light-curve In addition the absenceof any variability in the XMM-Newton RGS and OM light-curves related to the 11 h signal (seen in the Swift UV light-curve) has the potential to rule out the modulation at thisshort period Therefore we assume that the 357 h is mostlikely the Porb of the binary

Ignoring this long-term variability seen in the low fre-quency part of XMM-Newton periodogram we find signifi-cant signals at ω ≃ 177times10minus3 sminus1 in both (XMM-Newton

and Chandra) light-curves More precisely the periodicitieswe found are at 56504plusmn 068 s for the Chandra HRC light-curve and 56496plusmn 050 s for the XMM-Newton RGS light-curve The SwiftXRT observations are almost all too shortto usefully constrain the modulation at this short period

Figure 6 Lomb Scargle periodograms of the XMM-NewtonRGS1 (top) and the Chandra HRC (bottom) light-curves Vari-ous frequencies are indicated where Ω = 1Porb and ω = 1Pspin

(see text for more details) We present zoom-in plots around the177times10minus3 sminus1 in each periodogram

Figure 7 Top from top to bottom soft RGS1 light-curve hard

RGS1 light-curve and the hardness ratio The energy bandsfor the soft and hard light-curves are 235 ndash 370 A (03325 ndash

0528 keV) and 150 ndash 235 A (0528 ndash 0827 keV) respectively Bot-tom same as top but all folded at the 565 s spin period andnormalized to their respective mean rates Therefore the y-axisrepresents the relative amplitude of the modulation in both thesoft and hard bands

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 8: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

6 Aydi et al

Figure 3 First panel the XMM-Newton EPICpn X-ray light-curve Second panel the RGS1 X-ray light-curve The X-ray light-curvesshow two broad (sim 2 ndash 3 ks wide) dips separated by 126 ks consistent with the 357 h modulation seen in the Swift-UVOT data (seeSection 34) Third panel the uvw1 OM fast mode (black) and imaging mode (red) light-curves The first OM exposure is a grismspectrum hence the light-curves only start at sim 5 ks The blue line represents a best fit sine curve to the OM data with a period of376plusmn03 h This period is consistent within uncertainty with the period seen in the Swift-UVOT data (see Section 34 for further

discussion on the timing analysis) The y-axis in all the panels are counts per second

32 The Swift UVOT light-curve

The UVOT uvw2 light-curve (Fig 2) declined slowly afterday 150 before flattening off at around a uvw2 magnitude of142 by around day 200 with signs of a 01mag variabilitysuperimposed on the overall decline We find periodicities inthe light-curve which we discuss in Section 34 The UV datarebrightened after day 300 to reach a uvw2 magnitude of137 around day 320 Then the brightness started to declineagain reaching a uvw2 magnitude of 149 at day sim 385Fig B2 shows a direct comparison between the Swift X-raySwift UV and SMARTS optical light-curves

33 The XMM-Newton and Chandra UV and

X-ray light-curves

The XMM-Newton RGS1 X-ray light-curve the EPICpnX-ray light-curve and the OM UV fast and imaging modeslight-curves are all presented in Fig 3 The XMM-Newton

RGS1 light-curve shows evidence of variability on differenttimescales This includes two broad (sim 2 ndash 3 ks wide) dipsseparated by sim 126 ks sim 35 h The OM fast mode UV light-curve shows a variation characterized by a 376plusmn03 h periodwhich is also consistent (within uncertainty) with the dura-tion between the dips in the RGS1 light-curve Howeverno other periodicity is found in the OM light-curve TheChandra HRC light-curve (Fig 4) also shows evidence forshort-term variability which we discuss in Section 34

Figure 4 The Chandra High Resolution Camera (HRC) X-ray

light-curve The HRC energy range is 006 ndash 10 keV There is aclear short-term variability in the light-curve and it is discussed

in Section 34

34 Timing analysis

Since the SwiftUVOT uvw2 light-curve (Fig 2) shows evi-dence for intrinsic variability superimposed on the long termdecline we searched for periodicities in the data Fig 5shows a Lomb-Scargle periodogram (LSP) of the barycen-tric corrected uvw2 light-curve after subtracting a third or-der polynomial to remove the long term trend The mostsignificant peaks occur at periods of 35731 plusmn 00014 h and

nova V407 Lup 7

Figure 5 Upper panel the Lomb Scargle periodogram of thedetrended uvw2 light-curve revealing significant modulation atfrequencies of 67168 dminus1 (777times10minus5 sminus1 P = 3573 h) and217696 dminus1 (252 times 10minus4 sminus1 P = 1102 h with estimated un-certainties of 00003 dminus1) marked with red dashed lines Lowerpanel periodogram of the observing window showing Swift rsquos or-

bital frequency (dotted line)

11024plusmn 00002 h which are aliases of each other caused bySwift rsquos 16 h orbit One of these periods might represent theorbital period of the binary (Porb) The low cadence of theUVOT data means we cannot break the degeneracy betweenthe periods (the high-cadence observation interval betweendays 307 and 311 failed to resolve this issue) The amplitudeof the modulation is approximately 01mag

We performed a similar Lomb-Scargle analysis of theXMM-Newton RGS1 and Chandra HRC data (see Fig 6)The light-curve of the former shows considerable flux vari-ations while the flux in the latter is approximately con-stant over the duration of the observation The XMM-

Newton periodogram is dominated by a low frequency ofΩ = 777times10minus5 sminus1 (357 h) This signal only covers sim twocycles of the 357 h period and cannot be believed in iso-lation However as this signal is consistent with the longperiod seen in the UVOT data we identify it as the same357 h modulation

The long-timescale modulation and variability seen inthe XMM-Newton light-curves the time between the dipsand the weak modulation observed in the OM light-curve(see Figs 3 and 6) are all consistent with the 357 h signalseen in the Swift UV light-curve In addition the absenceof any variability in the XMM-Newton RGS and OM light-curves related to the 11 h signal (seen in the Swift UV light-curve) has the potential to rule out the modulation at thisshort period Therefore we assume that the 357 h is mostlikely the Porb of the binary

Ignoring this long-term variability seen in the low fre-quency part of XMM-Newton periodogram we find signifi-cant signals at ω ≃ 177times10minus3 sminus1 in both (XMM-Newton

and Chandra) light-curves More precisely the periodicitieswe found are at 56504plusmn 068 s for the Chandra HRC light-curve and 56496plusmn 050 s for the XMM-Newton RGS light-curve The SwiftXRT observations are almost all too shortto usefully constrain the modulation at this short period

Figure 6 Lomb Scargle periodograms of the XMM-NewtonRGS1 (top) and the Chandra HRC (bottom) light-curves Vari-ous frequencies are indicated where Ω = 1Porb and ω = 1Pspin

(see text for more details) We present zoom-in plots around the177times10minus3 sminus1 in each periodogram

Figure 7 Top from top to bottom soft RGS1 light-curve hard

RGS1 light-curve and the hardness ratio The energy bandsfor the soft and hard light-curves are 235 ndash 370 A (03325 ndash

0528 keV) and 150 ndash 235 A (0528 ndash 0827 keV) respectively Bot-tom same as top but all folded at the 565 s spin period andnormalized to their respective mean rates Therefore the y-axisrepresents the relative amplitude of the modulation in both thesoft and hard bands

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 9: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 7

Figure 5 Upper panel the Lomb Scargle periodogram of thedetrended uvw2 light-curve revealing significant modulation atfrequencies of 67168 dminus1 (777times10minus5 sminus1 P = 3573 h) and217696 dminus1 (252 times 10minus4 sminus1 P = 1102 h with estimated un-certainties of 00003 dminus1) marked with red dashed lines Lowerpanel periodogram of the observing window showing Swift rsquos or-

bital frequency (dotted line)

11024plusmn 00002 h which are aliases of each other caused bySwift rsquos 16 h orbit One of these periods might represent theorbital period of the binary (Porb) The low cadence of theUVOT data means we cannot break the degeneracy betweenthe periods (the high-cadence observation interval betweendays 307 and 311 failed to resolve this issue) The amplitudeof the modulation is approximately 01mag

We performed a similar Lomb-Scargle analysis of theXMM-Newton RGS1 and Chandra HRC data (see Fig 6)The light-curve of the former shows considerable flux vari-ations while the flux in the latter is approximately con-stant over the duration of the observation The XMM-

Newton periodogram is dominated by a low frequency ofΩ = 777times10minus5 sminus1 (357 h) This signal only covers sim twocycles of the 357 h period and cannot be believed in iso-lation However as this signal is consistent with the longperiod seen in the UVOT data we identify it as the same357 h modulation

The long-timescale modulation and variability seen inthe XMM-Newton light-curves the time between the dipsand the weak modulation observed in the OM light-curve(see Figs 3 and 6) are all consistent with the 357 h signalseen in the Swift UV light-curve In addition the absenceof any variability in the XMM-Newton RGS and OM light-curves related to the 11 h signal (seen in the Swift UV light-curve) has the potential to rule out the modulation at thisshort period Therefore we assume that the 357 h is mostlikely the Porb of the binary

Ignoring this long-term variability seen in the low fre-quency part of XMM-Newton periodogram we find signifi-cant signals at ω ≃ 177times10minus3 sminus1 in both (XMM-Newton

and Chandra) light-curves More precisely the periodicitieswe found are at 56504plusmn 068 s for the Chandra HRC light-curve and 56496plusmn 050 s for the XMM-Newton RGS light-curve The SwiftXRT observations are almost all too shortto usefully constrain the modulation at this short period

Figure 6 Lomb Scargle periodograms of the XMM-NewtonRGS1 (top) and the Chandra HRC (bottom) light-curves Vari-ous frequencies are indicated where Ω = 1Porb and ω = 1Pspin

(see text for more details) We present zoom-in plots around the177times10minus3 sminus1 in each periodogram

Figure 7 Top from top to bottom soft RGS1 light-curve hard

RGS1 light-curve and the hardness ratio The energy bandsfor the soft and hard light-curves are 235 ndash 370 A (03325 ndash

0528 keV) and 150 ndash 235 A (0528 ndash 0827 keV) respectively Bot-tom same as top but all folded at the 565 s spin period andnormalized to their respective mean rates Therefore the y-axisrepresents the relative amplitude of the modulation in both thesoft and hard bands

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 10: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

8 Aydi et al

Figure 8 From top to bottom soft Chandra HRC light-curvehard Chandra HRC light-curve and the hardness ratio all foldedover the spin period and normalized to their respective meanrates Therefore the y-axis represents the relative amplitude of

the modulation in both the soft and hard bands The energybands for the soft and hard light-curves are 30 ndash 500 A (0248 ndash

0413 keV) and 150 ndash 300 A (0413 ndash 0827 keV) respectively

Figure 9 Lomb Scargle periodograms of the AAVSO data takenbetween days 327 and 350 post-eruption The red dashed linerepresents the frequency ω = 177times10minus3 sminus1 (1Pspin) and theblue dotted line represents the sideband frequency (ω minus Ω)

The 565 s X-ray period and the longer 357 h UVX-ray period are clearly reminiscent of the typical spin period(Pspin) and Porb seen in IP CVs (Warner 1995) This impliesthat the system may host a magnetized WD The ratio ofPspinPorb is sim 044 typical for IPs (Warner 1995 NortonWynn amp Somerscales 2004) Therefore we suggest that thesim 565 s period is the Pspin of the WD

While the Chandra HRC power spectrum (Fig 6) isrelatively easy to read with a single dominant peak (at177times10minus3 sminus1 Pspin = 565 s) the XMM-Newton data showsa more complicated periodogram with some asymmetry inpeak amplitudes This suggests rather complicated signalbehaviour The different peaks around the central one (ω≃177times10minus3 sminus1) coincide with ω plusmn Ω and ω plusmn 2Ω side-bands Such signals are commonly seen in IPs (see eg Nor-

Figure 10 The AAVSO optical light-curve folded over the WD

spin period of 565 s using the same ephemeris used to fold theXMM-Newton RGS1 light-curves (Fig 7) The y axis representsthe change in magnitude relative to the mean brightness level(positive values mean brighter and negative values mean fainter)

ton Beardmore amp Taylor 1996 Ferrario amp Wickramasinghe1999)

Fig 7 represents the XMM-Newton RGS1 soft and hardlight-curves and the hardness ratio When folded at the Pspin

= 565 s the light-curves show near sinusoidal variation Suchvariation is expected from an IP (see eg Osborne 1988Norton amp Watson 1989 Norton et al 1992ba Norton 1993Warner 1995) however the RGS data were taken during thepost-nova SSS phase and therefore the mechanism respon-sible for such modulation might be different from that seenusually in IPs (see Section 54 for further discussion)

Fig 8 represents the folded Chandra HRC soft and hardlight-curves over the Pspin as well as the hardness ratio TheHRC light-curve (taken 340 days after the eruption) showstronger modulation in the harder bands while the hardnessratio is slightly modulated This might also suggest that thevariation over the Pspin is not simply due to absorption byan accretion curtain as it is usually the case for IPs (Warner1995)

Typically the physical reason behind an opticalX-raymodulation over the Pspin of theWD seen in IPs is attributedto the variation of the viewing aspect of the accretion curtainas it converges towards the WD surface near the magneticpoles or due to the absorption caused by the accretion cur-tains Therefore such a modulation is usually a sign of accre-tion impacting the WD surface near the magnetic poles andpossibly in this case a sign of accretion restoration Howeverthe XMM-Newton observations were taken during the SSSphase and therefore the soft X-rays are dominated by emis-sion from the H burning on the surface of the WD Thusthere must be another explanation for the Pspin modulationseen in the X-ray light-curves (see Section 54 for furtherdiscussion about the accretion resumption and the origin ofthe X-ray modulation)

We also performed Lomb-Scargle analysis of the opti-cal AAVSO data (unfiltered reduced to V band) obtainedbetween 327 and 350 days post-eruption The observationswere performed almost every other night and they consist of

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 11: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 9

Figure 11 The SALT HRS high-resolution spectra of days 164 and 209 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A (right)

with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0 Wepresent a zoomed in plot on the range between 5100 A and 5450 A to show the possible presence of weak emission lines in this rangeBelow 4200 A instrumental noise dominates the spectra

multiple exposures separated by sim40 s and spanning for sim4 h The power spectrum shows a peak at sim 177times10minus3 sminus1

the same as that seen in the Chandra power spectrum andwhich is an indication of the Pspin of the WD (Fig 9) Thesignals seen in the periodogram (Fig 9) at small frequenciesare not consistent with either the 357 h or the 11 h signalsseen in the UVOT light-curve and are most likely due to rednoise The spin phase-folded AAVSO optical light-curve isshown in Fig 10 We folded the light-curve using the sameephemeris used to fold the RGS1 light-curves While theAAVSO optical and XMM-Newton X-ray light-curves areout of phase caution is required when comparing these twolight-curves as the XMM-Newton observations were donemore than 150 days prior to the AAVSO ones (see Section 54for further discussion)

4 SPECTROSCOPIC RESULTS AND

ANALYSIS

41 Optical spectroscopy

411 Line identification

The strongest emission features in the first two SALT HRSspectra and the XMM-Newton OM grism spectrum on days164 168 and 209 (Fig 11 and 15) are the forbidden oxy-gen lines [O iii] 4363 A 4959 A 5007 A and [O ii] 732030 Aalong with the Balmer lines The lines are very broad withflat-topped jagged profiles Forbidden neon lines are alsopresent ([Ne iii] 3698 A blended with Hǫ and [Ne iv] 4721 Awhich might be blended with other neighbouring lines)Other forbidden neon lines such as [Ne iii] 3343 A 3869 A[Nev] 3346 A and 3426 A are also present in the SOAR spec-trum of day 269 (Fig 14) and in the Swift UVOT spectra(Section 42) Also present are weak permitted lines of he-lium (He i 4713 A 5876 A He ii 4686 A 5412 A and 8237 A -

the lines of the Pickering series at 4860 A and 6560 A mightbe blended with Hβ and Hα respectively) high excitationoxygen lines (Ovi 5290 A) and weak [O i] lines at 6300 Aand 6364 A The forbidden [N ii] line at 5755 A is strongand broad while the permitted N iii line at 4517 A is rela-tively weak and the one at 4638 A is possibly blended withother lines The [N ii] doublet at 6548 and 6584 A mightbe blended with broad Hα Relatively weak high ionizationcoronal lines of iron might also be present ([Fevii] 6087 A[Fexi] 7892 A and possibly [Fe ii] 5159 A and [Fe iii] 4702A)The spectra also show lines with a FWHM of sim 100 km sminus1

at sim 4498 A 5290 A and 7716 A (see Section 4135)

The optical spectra show three different type of linescharacterized by significantly different widths therefore inthe remainder of the paper we will use the following terms

bull ldquoVery narrow linesrdquo to denote those lines with a FWHMsim 100 km sminus1 (see Section 4135)

bull ldquoModerately narrow linesrdquo to denote those lines with aFWHM sim 450 km sminus1 such as He ii 4686 A (seeSections 4133 and 4134)

bull Broad lines to denote those lines with a FWHM sim3000 km sminus1 that originate from the ejecta such asthe Balmer and [O iii] lines (see Sections 4131and 4132)

From day 250 to day 317 the SALT HRS spectra arestill dominated by the broad forbidden oxygen lines (seeFigs 12 and 13) The Balmer lines are fading gradually whileother lines such as forbidden Ne and Fe start to disappearAt day 250 a ldquomoderately narrowrdquo and sharp He ii line at4686 A emerges and is accompanied by less prominent He iilines at 4200 A 4542 A and 5412 A Similar features of Ovi

5290 A and 6200 A emerge simultaneously and become moreprominent sim 30 days later On top of these two Ovi linesthe aforementioned ldquovery narrow linesrdquo are still present We

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 12: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

10 Aydi et al

Figure 12 The SALT HRS high-resolution spectra of days 250 263 and 279 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0

Below 4200 A instrumental noise dominates the spectra

Figure 13 The SALT HRS high-resolution spectra of days 285 304 309 and 317 plotted between 3900 ndash 5500 A (left) and 5450 ndash 8800 A

(right) with the flux in arbitrary units For clarity the spectra are shifted vertically The numbers in square brackets are days since t0Below 4200 A instrumental noise dominates the spectra

also detect similar emission features at sim 4498 and 6050 Athat we could not identify

All the ldquovery narrowrdquo and ldquomoderately narrowrdquo linesshow changes in radial velocity and structure unlike thebroad lines At day 285 He ii 4686 A becomes as strong as[O iii] 4363 A and surpasses it aftersim 300 days At this stageldquomoderately narrowrdquo Balmer features emerge superimposedon the pre-existing broad features

Further to the blue (below 4200 A) the SOAR medium-resolution spectrum of day 269 (Fig 14) shows ldquovery nar-rowrdquo emissions of Ovi at 3811 A and sim 4157 A Relatively

strong and broad lines of [Nev] at 3426 A and 3346 A and[Ne iii] at 3869 A and 3968 A are also present

The late SOAR medium-resolution spectra of days 484and 485 were of limited-range centred near to He ii at4686 A which appears stronger than Hβ (Fig 16) The lat-ter has a significant broad base In Table C2 we list the lineidentifications along with the FWHM Equivalent Widths(EWs) and fluxes of those lines for which an estimate waspossible

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 13: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 11

Figure 14 The SOAR medium-resolution spectrum of day 270 plotted between 3200 ndash 5800 A (left) and 5600 ndash 6800 A (right) with the

flux in erg cmminus2 sminus1 A1

Figure 15 The OM (Optical Monitor on board XMM-Newton)

grism spectrum plotted between 3000 and 6000 A The numbersin square brackets are days since t0

412 Spectral classification and evolution

Izzo et al (2018) obtained spectra as early as day 5 post-eruption These spectra show characteristics of the opticallythin HeN spectral class with ejecta velocity sim 2000 km sminus1They also point out the presence of Fe ii lines which theyattribute to a very rapid iron-curtain phase

The HRS spectra taken on days 164 and 209 were en-tirely emission lines and dominated by broad nebular forbid-den oxygen lines along with Balmer and forbidden neon andiron lines The spectra show that the nova is well into thenebular phase by then We expect that this phase startedaround a month from the eruption based on the fast light-curve evolution consistent with the spectra of Izzo et al

Figure 16 The SOAR medium-resolution spectra taken on day484 and day 485 plotted between 4300 and 5000 A with the fluxin arbitrary units

(2018) The presence of neon lines in the spectrum also sug-gest that V407 Lup might be a ldquoneon novardquo showing thatthe eruption occurred on a ONe WD which was confirmedby Izzo et al (2018) after deriving a Ne abundance sim 14times Solar The highlight of that study was the detectionof the 7Be ii 3130 A doublet and that V407 Lup an ONenova has produced a considerable amount of 7Be whichdecays later into Li They concluded that not only CO butalso ONe novae produce Li confirming that CNe are themain producers of Li of stellar origin in the Galaxy

The optical spectra show the presence of high ionizationcoronal lines (eg [Fevii] 6087 A and [Fexi] 7892 A) Suchlines can be attributed to photoionization from the central

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 14: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

12 Aydi et al

hot source during the post-eruption phase to a hot coronal-line-region physically separated from the ejecta responsiblefor the low ionization nebular lines or possibly to shockswithin the ejecta (see eg Shields amp Ferland 1978 Williamset al 1991 Wagner amp Depoy 1996 and references therein)

In the spectra of day 250 onwards ldquomoderately nar-rowrdquo lines of He ii and Ovi emerge the strongest beingHe ii 4686 A These lines show changes in their radial ve-locity between sim minus210 km sminus1 and 0 km sminus1 Similar narrowand moving lines have been observed in a few other novae(eg nova KT Eri U Sco LMC 2004a and 2009) and theirorigins have been debated (see eg Mason et al 2012 Mu-nari Mason amp Valisa 2014 Mason amp Munari 2014 and ref-erences therein see Sections 4133 and 4134 for furtherdiscussion)

413 Line profiles

4131 Balmer lines in the early spectra of nova V407Lup (days 5 8 and 11) the Balmer lines showed a FWHMof sim 3700 km sminus1 (Izzo et al 2016 2018) This FWHM haddecreased to sim 3000 km sminus1 by the first HRS spectra at day164 The lines then show a systematic narrowing with time(Fig 17) We measure the FWHM of these lines by applyingmultiple component Gaussian fitting in the Image Reductionand Analysis Facility (IRAF Tody 1986) and Python (scipypackages7) environments separately The values are given inTable C3

This line narrowing has been observed in many othernovae (see eg Della Valle et al 2002 Hatzidimitriou et al2007 Shore et al 2013 Darnley et al 2016) and can beattributed to the distribution of the velocity of the matterat the moment of ejection The density of the fastest mov-ing gas decreases faster than that of the slower moving gasleading to a decrease in its emissivity and in turn to the linenarrowing (Shore et al 1996)

While most novae occur in systems hosting a main-sequence secondary some nova systems have a giant sec-ondary For such systems an alternative explanation for theline narrowing has been presented by Bode amp Kahn (1985)after efforts to model the 1985 eruption of RS Oph Theseauthors suggested that shocks and interaction between thehigh velocity ejecta and low-velocity stellar wind from thecompanion (red giant in case of RS Oph) are responsible fordecelerating the ejecta which manifests as line narrowing

The Balmer lines have flat-topped jagged profiles(probably due to clumpiness in the ejecta) with no changesin radial velocity indicating an origin associated with theexpanding ejecta At day 250 ldquomoderately narrowrdquo Balmeremission features (FWHMsim 500 km sminus1) superimposed onthe broad emission profiles start to emerge They becomeprominent in later spectra (day 304 onwards see Fig 17)These emission features show variability in structures andradial velocity (see Table C4) which is difficult to measureaccurately due to the contamination by the broad emissioncomponent However due to their width and the changein radial velocity it is very likely that these features origi-nate from the inner binary system possibly associated withan accretion region (as it is unreasonable to associate such

7 httpswwwscipyorg

Figure 17 The evolution of the line profiles of the Balmer emis-

sion From left to right Hα Hβ Hγ and Hδ From top to bottomdays 164 209 250 263 279 285 304 309 and 317 A heliocen-tric correction has been applied to the radial velocities The fluxis in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1) The top-left panelshows an example of multiple Gaussian fitting that was used toderive the FWHM

ldquomoderately narrowrdquo features which show such a change inradial velocity with emission from the expanding ejecta)

At this stage Hγ is still completely dominated by the[O iii] line at 4363 A However ldquomoderately narrowrdquo Hδemission became prominent while its broad base has com-pletely faded We also note a dip or absorption feature to theblue of Hβ Hγ and Hδ This dip becomes very prominentin the last spectrum at sim minus650 km sminus1 (Fig 17)

These ldquomoderately narrowrdquo Balmer features becameprominent sim 30 days after the emergence of the narrow He iilines (see Section 4133) This is expected because suchfeatures which may originate from the inner binary systemcan only be seen clearly once the broad Balmer emission hasweakened significantly They also show single-peak emissionin most of the spectra indicating that if they are originatingfrom the accretion disk the system is possibly to have a lowinclination (eg close to face-on between 0 and 15 see fig239 in Warner 1995)

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 15: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 13

Figure 18 The evolution of the profiles of the ldquomoderately narrowrdquo emission lines From left to right He ii 4542 A He ii 4686 A He ii5412 A Ovi 5290 A and Ovi 6200 A From top to bottom days 250 263 279 285 304 309 and 317 A heliocentric correction is appliedto the radial velocities The flux is in arbitrary units The red dashed lines represent the rest-wavelength of each line (Vrad = 0km sminus1)

4132 Forbidden oxygen lines broad nebular emis-sion features of forbidden oxygen dominated the late spec-tra of nova V407 Lup (day 164 onwards) The [O iii] 4363 Ais superimposed over Hγ hence the blue-shifted pedestalfeature The forbidden oxygen emission features are broad(FWHM sim 3000 km sminus1) and show jagged profiles possiblyan indication of clumpiness in the ejecta

We measured the FWHM of the [O iii] 4363 A by apply-ing multiple component Gaussian fitting in the IRAF andPython environments separately The values are listed in Ta-ble C5 The other [O iii] lines are blended with other lines orwith each other Similar to the Balmer lines the [O iii] linesshow a systematic narrowing with time (Section 4131) butno changes in radial velocity

4133 The He II moderately narrow lines severalnovae have shown relatively narrow He components superim-posed on a broad pedestal during the early nebular stagesWhile the spectra evolve the narrow components becomenarrower and stronger (eg nova KT Eri U Sco LMC 2004aand 2009) For these novae it has been suggested that duringthe early nebular stage when the ejecta are still opticallythick the relatively narrow components originate from anequatorial ring while the broad pedestal components origi-nate from polar caps (see eg Munari et al 2011 MunariMason amp Valisa 2014 and references therein) Then whenthe ejecta become optically thin these components becomenarrower by a factor of around 2 (FWHM changing from

Figure 19 A sample of the complex He ii 4686 A ldquomoderatelynarrowrdquo line from day 309 The line is possibly a complex of

3 components the green line represents the observation whilethe blue solid line is the total fit of the 3 Gaussian components

(dashed lines)

sim 1000 km sminus1 to sim 400 km sminus1) and show cyclic changesin radial velocity Such cyclic changes in radial velocity canonly be associated with the binary system most likely theaccretion disk (Munari Mason amp Valisa 2014)

This is not the case for V407 Lup as only at day 250 the

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 16: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

14 Aydi et al

ldquomoderately narrow linesrdquo of He ii start to emerge (it is pos-sible that early in the nebular stage these lines have beenblended and dominated by the broad and prominent linescoming from the ejecta) We detect He ii 4686 A 4542 A5412 A and 4200 A The latter is heavily affected by theincreased noise at the edge of the blue arm of HRS Theprofiles of the three higher SN He ii lines are presented inFig 18 We measure the radial velocity and FWHM of theHe ii lines by fitting a single Gaussian component The mea-surements are illustrated in Table C6 The radial velocities ofthe He ii lines range between sim minus210 km sminus1 and 0 km sminus1

and they have an average FWHM of sim 450 km sminus1 Thisrange of velocities indicates that the system might have anegative systemic velocity of around minus100 km sminus1 The threelines show consistent velocity and structure changes acrossthe different spectra suggesting that they are originatingfrom the same regions

The He ii ldquomoderately narrow linesrdquo are complex andcan be subdivided into at least three components (1)a medium width component (MWC) with FWHM sim300 km sminus1 (2) a small width component (SWC) withFWHM sim 100 km sminus1 blended with the MWC in most ofthe spectra (3) and a broad base component (BBC Fig 19)Such complex structures of He ii lines are characteristic ofmCVs and they are variously associated with emission fromthe accretion stream the flow through the magnetosphereand the secondary star (Rosen Mason amp Cordova 1987Warner 1995 Schwope Mantel amp Horne 1997) We alsonote the presence of a very weak red-shifted emission at sim+600 km sminus1 from He ii 4686 A that strengthens and weak-ens from one spectrum to another (Fig 18)

We measure the velocity of the different components ofthe He ii 4686 A line by applying multiple Gaussian com-ponent fitting (Fig 19) The radial velocity and FWHM ofthe MWC and SWC of the He ii 4686 A line are listed in Ta-bles C7 and C8 respectively Although the two componentsare clearly out of phase their velocity measurements shouldbe regarded as uncertain and caution is required when inter-preting them since it is not straightforward to deconvolvethe different components We also measure the full width ofthe BBC (see Table C9)

Although the complexity of the He ii line profiles makesit difficult to draw conclusions about the origins of the dif-ferent components we suggest that the SWC characterizedby a FWHM sim 100 km sminus1 must originate from an area oflow velocity (possibly the heated surface of the secondary)However the other two components are possibly associatedwith emission from an accretion region It is possible thata fourth component is also present in the He ii lines whichadds to the complexity

4134 The O VI moderately narrow lines fromday 250 we detect relatively weak ldquomoderately narrowlinesrdquo of Ovi 5290 A and 6200 A These two lines are ini-tially dominated by ldquovery narrow linesrdquo (see Fig 18 andSection 4135) possibly associated with a different elementand certainly originating from a different region At day 304the intensity of the Ovi lines becomes comparable to theneighbouring ldquovery narrow linesrdquo and they form a blend Inaddition we detect a similar line at 6065 A (possibly N ii)We derive the radial velocity of the ldquomoderately narrowrdquo

Figure 20 The evolution of the profiles of the three ldquovery narrowlinesrdquo (VNL) plotted against wavelength From left to right Ne iv

4498 A Ovi 5290 A and Ne iv 7716 A From top to bottom days164 209 250 263 279 285 304 309 and 317 At day 285 theldquovery narrow linesrdquo stand out very clearly from the neighbouringldquomoderately narrow linesrdquo (MNL) as indicated on the plot Aheliocentric correction is applied to the radial velocities

Ovi lines by fitting a single Gaussian The values are listedin Table C10

There is no clear correlation between the velocity andthe structure of the ldquomoderately narrowrdquo Ovi lines withthose of their He ii counterparts It is worth noting that theionization potential of He ii is sim 54 eV while that of Ovi issim 138 eV

4135 Very narrow lines a remarkable aspect of theoptical HRS spectra is the presence of very narrow movinglines These ldquovery narrow linesrdquo have an average FWHM ofsim 100 km sminus1 and show variation in their velocity (rangingbetween minus150 and +20 km sminus1) width and intensity Themost prominent lines are at sim 5290 A (possibly Ovi 5290 A)and sim 7716 A (possibly Ne iv 77159 A) A less prominentline is at sim 4998 A (possibly Ne iv 44984 A) The line iden-

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 17: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 15

tification is done using the CMFGEN atomic data8 Theselines stand out in the first few spectra while a broader neigh-bouring component emerges later so they form a blend InFig 20 we present the evolution of the lines

We measured the radial velocity and width of these linesby applying single Gaussian fitting The values are listed inTable C11 Their width associates them with a region oflow expansion velocity If these are indeed high ionizationOvi and Ne iv lines they must originate from a very hotregion Belle et al (2003) found similar lines of Nv in thespectra of the IP EX Hydrae and they have attributed theseto an emission region close to the surface of the WD Notto rule out that disk wind might also be responsible forthe formation of such lines (see eg Matthews et al 2015Darnley et al 2017)

414 Radial velocities

Despite detecting changes in radial velocity of the He ii andOvi lines we failed to derive any periodicity from their ra-dial velocities This is expected as the SALT spectra aretaken on different nights separated by a few days up toa month With such a cadence we would not expect tofind modulation in the radial velocity curves Phase-resolvedspectroscopy is needed to do this and to investigate thestructure of the system via Doppler tomography (see egKotze Potter amp McBride 2016) In addition the emissionfrom different components (such as the accretion streamdisk and curtain and the secondary) adds to the complex-ity of the line profiles

The absence of any spin-period-related modulation inthe velocity curves is also expected due to the long exposuretime and cadence of the spectra The exposure time of theSALT HRS spectra is more than three times that of thePspin thus any possible WD spin-dependant effects will besmeared out in the spectra

Fig 21 shows the phase-folded ( over the Porb) radialvelocities of the MWC and SWC of the He ii 4686 A Thevelocities of these two components are anti-correlated TheSWC and MWC are out of phase and they show oppositeradial velocity curves neither of which is consistent with the357 h period The opposite phase is probably an indicationfor the origin of these components where the SWC might beassociated with emission from the surface of the secondaryand the MWC is originating from an accretion region aroundthe WD (accretion disk see eg Rosen Mason amp Cordova1987 Schwope Mantel amp Horne 1997) However the lackof periodicity in the velocity curves related to the orbitalperiod makes it difficult to confirm this claim

The radial velocity amplitude of the emission lines fromCVs is strongly dependent on the inclination of the systemas well as the Porb and the mass ratio Ferrario Wickramas-inghe amp King (1993) have modelled IPs with a truncatedaccretion disk and a dipolar magnetic field resulting in twoaccretion curtains above and below the orbital plane whichare responsible for most of the line emission For such sys-tems the amplitude of the radial velocities can range fromsim 200 km sminus1 up to 1000 km sminus1 depending mainly on the

8 httpkookaburraphyastpitteduhillierwebCMFGEN

htm

Figure 21 Radial velocities of the He ii 4686 A medium widthcomponent (MWC blue circles) and small width component

(SWC red squares) plotted against the orbital phase The evolu-tion of the velocities is anti-correlated for the two components

inclination of the system They also showed that velocitycancellation can occur if both curtains are visible leadingto velocity amplitudes in the range of sim 200 ndash 300 km sminus1This happen in the case of either a low-inclination systemwhere both curtains are visible through the centre of thetruncated disk or if the system is eclipsing (edge-on) Theamplitude of the radial velocities derived from the spectralemission lines of V407 Lup is around 200 km sminus1 This sug-gests that the system is possibly at a low inclination

415 Sodium D absorption lines

The sodium D absorption lines D1 (5896 A) and D2(5890 A) are well-known tracers of interstellar material (seeeg Munari amp Zwitter 1997 Poznanski Prochaska amp Bloom2012) In some cases circumstellar material around the sys-tem can also contribute to Na i D absorption For recurrentnovae this material has been proposed to be due to ejectafrom previous eruptions

In the high-resolution spectra the Na iD lines are a com-plex of at least two components both blue-shifted (Fig 22)There is a relatively weak component at sim minus50plusmn 3 km sminus1

and a more prominent one at sim minus8plusmn 3 km sminus1 Takinginto account a presumable systemic velocity of the system(sim minus100 km sminus1) both components would be red-shiftedWe measure a FWHM of sim 35 km sminus1 for each componentWe measure an EW of sim 094plusmn002 A for the D1 line and sim063plusmn002 A for the D2 line These values are used to derivethe reddening in Section 511

42 Swift UVOT spectroscopy

The UVOT grism spectra (Fig 23) are dominated by broademission lines of forbidden O and Ne along with H Balmerand Mg ii resonance lines A list of the observed lines andidentifications is given in Table C12 Since the Swift UVOTspectra from the grism exposures have an uncertainty inthe position of the wavelengths on the image the individualspectra were shifted to match the [Nev] 3346 A and 3426 A

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 18: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

16 Aydi et al

Figure 22 The profiles of the Na i D1 (top) and D2 (bottom)absorption lines Each colour corresponds to a different observa-tion At least two main components can be identified in the linesThe blue dashed line represents the average radial velocity of the

weaker component The magenta dotted line represents the aver-age radial velocity of the stronger component The red solid line

represents the null radial velocity A heliocentric correction hasbeen applied to the radial velocities

lines The shifts were at most 15 A Any intrinsic shift of theline spectrum is thus undetermined The emission seen at1759 A and 1855 A in the first order spectrum is uncertaindue to the low SN However in the grism image we cansee the 1750 A line in the second order spectrum which liesalongside the first order The 1860 A line in the second orderis affected by the wings of a nearby first order line andcannot be confirmed that way There is no significant line ofO ii] at 2471 A suggesting that the higher ionization statedominates The He ii 2511 A line may be blended with anunidentified feature around 2533 A The line at sim 2978 Ais possibly Neviii 2977 A (see eg Werner Rauch amp Kruk2007) A likely identification of the broad blend at 4714 Ahas been given with the SALT spectrum (He i 4713 A seeSection 411) Inspection of the spectrum in Fig 23 showsthat the Ne and O lines are much stronger than the H and He

lines in this nova which is consistent with the ONe natureof the WD as concluded by Izzo et al (2018)

The flux and width of the stronger lines of Mg ii O iii[Nev] and [Ne iii] have been measured for each of the eightUV spectra and the values are shown in Table C13 Themethod used was to plot the line select the points at whichthe line merges into the background and then sum the fluxover the line minus the background The flux uncertaintyis around 15 so the forbidden line ratios are as expectedin the low density case The Full Width at Zero Intensity(FWZI) clearly depends on the strength of the line Weakerlines merge earlier with the noise We also see larger FWZIfor longer wavelengths up to sim 7000 km sminus1

43 Swift X-ray spectral evolution

The initial detection of an X-ray source once V407 Luphad emerged from behind the Sun (day 150) showed it tobe in the supersoft regime already at a consistently highflux (Fig 25) Therefore if there was a high-amplitude fluxvariability phase as seen in some well-monitored novae (egV458 Vul ndash Ness et al 2009 RS Oph ndash Osborne et al 2011Nova LMC 2009a ndash Bode et al 2016 Nova SMC 2016 ndash Aydiet al 2018) it occurred while V407 Lup was behind the Sunand hence unobservable to Swift

The early X-ray spectra could be acceptably well mod-elled below 1 keV with a TMAP9 absorbed plane parallelstatic Non-Local Thermal Equilibrium (NLTE) stellar at-mosphere model (Rauch et al 2010 grid 003 was used) Thetbabs absorption model (Wilms Allen amp McCray 2000)within XSPEC was applied using the Wilms abundancesand Verner cross-sections this parameter was allowed tovary in the range (115 ndash 312) times 1021 cmminus2 based on theanalysis of the XMM-Newton RGS and Chandra LETGspectra (Sections 44 and 45) A sample of the spectra ob-tained during the monitoring is shown in Fig 24 whileFig 25 shows the results of the modelling of the super-soft component The luminosity was derived using an as-sumed distance of 10 kpc which is currently unknown (seeSection 512)

We note however that an absorbed blackbody modelwith a Nvii edge at sim065 keV (see Section 44) provides astatistically better fit than the more physically-based atmo-sphere model For example considering the spectrum ob-tained on day 200 while a simple blackbody is a poor fitwith C-statdof = 61166 including the oxygen absorp-tion edge decreases this to C-statdof = 10965 the TMAPatmosphere grid leads to C-statdof = 53366 underesti-mating the observed X-ray flux between 08 ndash 1 keV Fit-ting the spectra with the blackbody+edge model there isno evidence of a temporal trend for the optical depth ofthe edge over time with τ typically lying in the range1 ndash 4 The absorbing column required for such blackbodyfits is higher than for the atmosphere parameterisation atsim 3times 1021 cmminus2 While the blackbody+edge model is statis-tically preferred the luminosities from these fits are around

9 TMAP Tubingen NLTE Model Atmosphere Pack-agehttpastrouni-tuebingende~rauchTMAFflux_

HHeCNONeMgSiS_genhtml

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 19: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 17

Figure 23 The Swift UV-optical grism spectra plotted between 1700 and 5100 A The blue spectrum represents the first four spectracombined (from 2017 March 7 to 14) and the black spectrum represents the second four spectra combined (from 2017 March 26 to April23)

Figure 24 The top panel shows a sample of Swift X-ray instru-mental spectra from days 1501 2548 3351 3517 and 3840since eruption (t0) Note that the FWHM of the energy resolution

at 05 keV is sim 125 eV These sample spectra have been modelledwith a combination of an atmosphere (as considered throughout

the rest of the paper) and a fainter optically-thin component toaccount for the higher energy photons The lower panel shows theresiduals plotted in terms of sigma

a factor of ten higher than those from the atmosphere gridsbeing sim a few times 1039 erg sminus1 between days 150 and 300

It is noticeable that the later spectra (day 325 onwards)show evidence for harder X-ray emission above sim 15 keVThis corresponds to the time at which the X-ray source had

faded sufficiently that PC mode could be used The WTmode suffers from a higher background level10 in the caseof the data collected for V407 Lup the WT backgroundbegins to dominate the source emission around 1 keVmeaning that any such harder component cannot be easilymeasured

Using the dataset collected on 2018 January 23 (day486) at which point the SSS emission has completely fadedaway this harder spectral component can be parameterizedas optically thin emission with kT gt 12 keV and a 2 ndash 10 keVluminosity of sim 4times1033 erg sminus1 assuming a very uncertaindistance of 10 kpc (Section 512) This is the order of magni-tude expected for the luminosity from an intermediate polarassuming a typical accretion rate for a CV with a period ofsim 3 hr of sim 10minus8 ndash 10minus7 M⊙ yrminus1 (Patterson 1994 Pretoriusamp Mukai 2014) Even at a shorter distance (d sim 3 ndash 5 kpc)the hard X-ray luminosity would still be in the right rangeexpected from IPs

Other novae observed by Swift have also shown harder(gt1 keV) emission ndash see Osborne (2015) as well as the sam-ple of novae discussed by Schwarz et al (2011) Howeverthese objects typically showed evidence for this harder com-ponent from early on in the outburst before the start of theSSS phase rather than rdquoswitching onrdquo part way through thenova evolution [although V959 Mon (Page et al 2013) didreveal a slow rise and fall of the 08 ndash 10 keV emission] In

10 httpwwwswiftacukanalysisxrtdigest calphptrail

shows the typical count rate per column

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 20: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

18 Aydi et al

Figure 25 Results from fitting the Swift XRT spectra (with TMAP atmosphere model see text for more details) plotted againstdays since t0 From top to bottom the X-ray count rate the temperature of the SSS emission the best-fit absorbing column densityconstrained to lie in the range of (115 ndash 312)times1021 cmminus2 from the fits to the Chandra and XMM-Newton grating spectra marked bythe horizontal dashed lines bolometric luminosity for which a distance of 10 kpc is assumed (Section 512 ndash this distance is likely anupper limit for what the actual distance might be) For comparison for a 125M⊙ WD the Eddington luminosity is 16 times1038 erg sminus1

these cases this harder component was explained as shockseither within the nova ejecta or with external material suchas the wind from a red giant secondary

Assuming the same spectral shape as above (opticallythin component with kT gt 12 keV) we can place a limiton the luminosity of this component during the WT ob-servations of up to a factor of ten below that measured atlate times (after day 325) This suggests that this harderX-ray component has turned on significantly after the novaoutburst and could therefore be a signature of restarting ac-cretion However with the available data it is not possibleto rule out that the emission may be caused by shocks

44 XMM-Newton high-resolution X-ray

spectroscopy

The absorbed flux at Earth as measured withthe RGS at day 168 over the 14 ndash 38 A range was(12plusmn05)times10minus9 erg cmminus2 sminus1 The RGS spectrum (Fig 26)is dominated by a bright SSS continuum with relativelyweak absorption lines covering most of the spectral rangeThe most obvious features are absorption edges from Nvii

at 186 A and O i 228 A and absorption lines from Nvii 1s-2p3p4p5p (rest-wavelengths at 2474 A 209 A 1983 Aand 1936 A) Ovii 1s-2p (rest-wavelength at 216 A) andNvi 1s-2p3p (rest-wavelengths at 2878 A and 249 A)

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 21: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 19

Figure 26 The XMM-Newton RGS spectrum (red) taken on day 1685 The spectrum is combined from RGS1 and RGS2 using theSAS tool rgsfluxer The Chandra spectrum (blue) taken on day 3406 is also plotted for comparison and has been multiplied by a

factor of 30 It is worth noting that the Chandra LETG spectral range is much larger (see text for more details)

These lines are shifted by at most minus400 km sminus1 In additionthe Ovii and Nvii 1s-2p lines contain a fast component ofminus3200 km sminus1 A line at 285 A may either be Cvi shiftedby minus400 km sminus1 or Nvi shifted by minus3200 km sminus1 Furtherwe find interstellar absorption lines of O i 1s-2p (235 A)and N i 1s-2p (313 A)

The continuum can be parameterized surprisingly wellby a blackbody fit plus an absorption edge at sim 186 A to re-produce the Nvii edge (τ=094 equivalent to a column den-sity 15times1017 cmminus2) yielding Teff=61times105 K (kT=53 eV)NH = 312times1021 cmminus2 (Fig 27) The abundances of theinterstellar O and N were reduced to 06 solar The col-umn density of the Nvii edge suggests that it arises from ahot plasma (the ionization potential of Nvii is sim 667 eV)The presence of such a deep absorption edge while seeingldquoshallowrdquo absorption lines could indicate that the plasma ishighly ionized A high degree of ionization makes the plasmamore transparent and possibly this can explain why the con-tinuum is best fitted by a blackbody model Note that theChandra spectrum on day 340 shows evidence for even fewerabsorption lines (see Section 45) The continuum normaliza-tion corresponds to a radius of 52times104 km (assuming spher-ical symmetry this is an order of magnitude larger than atypical WD radius) at an assumed distance of 10 kpc (Sec-

tion 512) While a bloated WD is a possible interpretationoverestimates of the radius are quite common when usingblackbody fits In Appendix A we present conversion equa-tions for the parameters that are distance dependent (egradii and luminosities) We also present a range of values forthese parameters using different distance assumptions

We also experimented with the NLTE atmosphere mod-els from TMAP (Rauch et al 2010 Fig 27) and with thesynthetic models for expanding atmospheres ldquowind-typerdquomodel (van Rossum 2012 Fig B3) Neither of these modelsreproduced the observed RGS spectrum of V407 Lup Theclosest approximation with a plane-parallel static NLTETMAP model yields Teff=98times105 K (kT=85 eV) NH =14times1021 cmminus2 and an absorption line velocity of 290 km sminus1

(see Fig 27) The absorbed and unabsorbed X-ray (15 ndash38 A 033 ndash 1KeV) fluxes derived from this model are 119times10minus9 erg cmminus2 sminus1 and 545 times10minus9 erg cmminus2 sminus1 respec-tively The X-ray flux derived from the TMAP model is morethan 95 the bolometric flux therefore this flux equates toan absolute bolometric luminosity of sim 65times1037 erg sminus1 ata distance of 10 kpc (see Section 512 and Appendix A) Wenote that this TMAPmodel also contains a few weak absorp-tion lines The comparison with the wind-type atmosphere

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 22: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

20 Aydi et al

Figure 27 The top panel shows a blackbody model (blue) including an edge at sim 186 A (to reproduce the Nvii edge see text formore details) a one-component TMAP model (orange) and a two-components TMAP model (red) all fitted to the XMM-Newton RGSspectrum (gray) The parameters of each model are given in the plot The TMAP models are plane parallel thus without a radius in the

model itself Therefore the radii shown in the legend of the TMAP models (1-T and 2-T) were derived assuming spherical symmetryThe bolometric luminosities in the legend were derived using the values of the radii and Teff via Stefan-Boltzmann law These differ fromthe luminosities derived in the text based on the X-ray fluxes (see text for more details) The abundances of the interstellar O and Nwere reduced to 06 solar in the blackbody model The bottom panel shows the residuals of the fits as they vary against wavelength Theluminosities and radii derived here are distance dependent and therefore a range of these values are presented in Appendix A

models suggests Teff between 55times105 K and 60times105 K(kT=47 ndash 52 eV)

Due to the complexity of the spectrum we consider theidea that the atmosphere is not homogeneous possibly dueto hotter emitting regions towards the poles Therefore wecarried out modelling using two TMAP model componentsfrom the grid 003 The parameters of the models are givenin Table C14 The combination of these two models resultsin a reasonably good fit to the continuum and it is presentedin Fig 27 The total flux derived from the two-componentTMAP models is 122times10minus9 erg cmminus2 sminus1 and the total un-absorbed flux is 930times10minus9 erg cmminus2 sminus1 At a distance of10 kpc (see Section 512 and Appendix A) this equates toan absolute bolometric luminosity of sim 11times1038 erg sminus1

45 Chandra high-resolution X-ray spectroscopy

The Chandra LETG spectrum taken on day 340 is presentedin Fig 26 together with the XMM-Newton RGS spectrumThe observed (absorbed) integrated flux measured with the

Chandra LETG is 289times10minus10 erg cmminus2 sminus1 which is a fac-tor of sim 45 lower than the observed (absorbed) integratedflux measured with XMM-Newton RGS on day 168 (see Sec-tion 44) It is worth noting that the RGS spectrum is sen-sitive to a shorter range (6 ndash 38 A) compared to the LETGspectrum (12 ndash 175 A ie a part of the flux is not includedin the RGS spectrum) Even if NH has not decreased thiswould imply that the effective temperature which scales asLminus14 has decreased by a factor of 26 This is can be ruledout by examining Fig 26 which does not show such a shift intemperature (the peaks of the RGS and LETG spectra showa decrease of temperature by a factor of 14) We concludethat we are possibly only observing a small portion of theWD surface This might be an indication that at this stage(sim day 340) accretion has resumed and the WD is partiallyhidden by an accretion disk andor accretion curtains (seeSection 54 for further discussion) or it may instead meanthat the temperature is not homogeneous on the WD surfaceand supersoft X-rays are emitted only in a restricted regionof the surface possibly on the polar caps (see eg ZemkoMukai amp Orio 2015 Zemko et al 2016) Another intriguing

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 23: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 21

Figure 28 The top panel shows Model 1 (red) fit to the ChandraLETG spectrum (black) The model consists of two components

a TMAP atmosphere model and a bvapec Collisional IonizationEquilibrium (CIE) plasma model The carbon abundance in theCIE model are enhanced to 196 times the solar abundance Seetext for more details The bottom panel shows the residuals ofthe fit as they vary against wavelength

element in the spectrum is that the Nvii absorption featureappears blue-shifted by sim minus4600 km sminus1 which is a largervelocity compared to the blue-shift measured on day 168 inthe RGS spectrum (Fig 26)

We tried two models to fit the Chandra LETG spec-trum In the first model (Model 1 parameters given inTable C15) we assumed that the bulk of the flux comesfrom the WD surface with an additional component ofplasma in collisional ionization equilibrium For this pur-pose we used a TMAP atmosphere model with a bvapec

model based on the Astrophysical Plasma Emission Code

(APEC Smith et al 2001) with enhanced carbon abun-dances (196 times solar abundance) and assuming solar abun-dances for the other neutral absorbing elements The fit isnot perfect and underestimates the observed flux by 10 sowe conclude that there must be an additional componentpossibly a plasma at a different temperature and electrondensity since the residual flux seems to be in emission linesWe experimented by varying the abundances or the wind ve-locity as red-shift but this did not result in a better fit Theemission lines appear to be broadened by sim 5000 km sminus1although we had to fix this value to avoid obtaining an un-reasonably high velocity Similarly to the RGS spectrumwe also cannot fit well the shallow absorption line profilesThe total absorbed and unabsorbed X-ray (in the range 15 ndash50 A 0248 ndash 0827KeV) fluxes derived from this model are274times10minus11 erg cmminus2 sminus1 and 476times10minus10 erg cmminus2 sminus1 TheX-ray flux derived fromModel 1 are more than 95 the bolo-metric flux therefore this equates to an absolute bolometricluminosity of sim 57times1036 erg sminus1 at 10 kpc (see Section 512and Appendix A) Fig 28 represents the fit of Model 1

In the second model (Model 2 parameters given in Ta-ble C16) we assumed that there is a homogeneous WDatmosphere and that the polar caps become heated inexcess because of accretion and emit as an additional

blackbody component Because a blackbody energy distri-bution is broader than just the atmospheric spectral en-ergy distribution this results in the assumption that thereis more absorption of soft flux partially lifting the dis-crepancy in NH value (see Section 511) The best fitin this case yields NH =187times1021 cmminus2 The absorbedflux coming from the blackbody component alone wouldbe 215 times 10minus11 erg cmminus2 sminus1 and an unabsorbed flux of653 times 10minus9 erg cmminus2 sminus1 At a distance of 10 kpc (see Sec-tion 512) this would imply an absolute X-ray luminos-ity of sim 78 times 1037 erg sminus1 (in the range 15 ndash 50 A 0248 ndash0827KeV) which seems absurd because it is four orders ofmagnitude higher compared to heated polar caps of mag-netic CVs (eg Zemko et al 2017 and references thereinsee also Section 54) It is worth noting that even at shorterdistances (3 ndash 5 kpc see Appendix A) this luminosity is stillthree orders of magnitude higher compared to heated polarcaps of magnetic CVs

We also experimented with ldquowind-typerdquo models of vanRossum (2012) for an expanding atmosphere to fit theChandra spectrum These did not result in a good fit Ata distance of 10 kpc the best fit parameters are Teff =45 times 105 K NH = 17 times 1021 cmminus2 and log g (logarithmof the surface gravity) = 824 Fig B3 shows a sample ofthe best fit models for various wind asymptotic velocitiesvinfin and mass-loss rates M

5 DISCUSSION

51 Reddening distance and eruption amplitude

511 Reddening

Estimating the distance to CNe is not straightforward butit is a crucial step in deriving the properties of the eruptionand its energetics Estimating the reddening is vital for es-tablishing the distance and we investigate this using variousmethods below

The reddening maps of Schlafly amp Finkbeiner (2011)indicate AV = 088 in the direction of nova 407 Lup (l =3301 b = 957) which should be regarded as an approxi-mate lower limit for any object that is sufficiently distantFrom the modelling of the X-ray grating spectra we derivedNH ranging between 14 and 312 times 1021 cmminus2 Using therelation between NH and AV from Zhu et al (2017) wederive AV between sim 067 and 150 (plusmn 002) Caution is re-quired when drawing conclusions from these values as theyare model dependent

Poznanski Prochaska amp Bloom (2012) presented em-pirical relations between the extinction in the Galaxy andthe EW of the Na i D absorption doublet Using theseand the measured EW of the Na i D absorption lines at59900 A (D1) and 58960 A (D2 see Section 415) we de-rive E(B minus V ) = 093plusmn 025 and AV = 289 plusmn 025

Using colours of novae around maximum has also beensuggested as a way to derive reddening (van den Berghamp Younger 1987) These authors derived a mean intrinsiccolour (BminusV )0 = +023 plusmn 006 for novae at maximum lightand (BminusV )0 = minus002 plusmn 004 at t2 We lack broadband ob-servations exactly at maximum and t2 however we measure(B minus V ) sim 08 plusmn 004 at tmax + 05 d using SMARTS Thispoints towards E(B minus V ) and AV values close to that de-

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 24: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

22 Aydi et al

rived from the Na i D EWs We consider this interpretationas uncertain due to the fast decline and rapid change of thecolours around maximum in the case of V407 Lup

The reddening values we derive from the optical spec-troscopy and photometry are much higher than the valuesderived from the X-ray modelling (Sections 44 and 45) andthe ones from the reddening maps (see above) This discrep-ancy suggests that either Poznanski Prochaska amp Bloom(2012) relations are unreliable or the nova is embedded in acloud of interstellar dust Note that the IP identity of V407Lup (see Section 55) also adds to the discrepancy betweenthe reddening derived from the optical and X-ray spectra Ifthe spin modulation seen in the X-ray data is indeed due toabsorption from the system (Section 54) this means thatthe X-ray total absorption is due to two components theinterstellar medium (ISM) and the system Therefore theabsorption due to the ISM is less than the total absorptionand the AV derived from the X-ray spectra might be evensmaller

Izzo et al (2018) derived an intrinsic extinction ofE(B minus V ) = 024 plusmn 002 and therefore AV = 074 for V407Lup in good agreement with the reddening we derive fromthe X-ray spectra These authors have used the diffuse in-terstellar bands as well as the van den Bergh amp Younger(1987) relations to derive the reddening The assumption ofdifferent tmax t2 and the use of the broadband photometryreported in Chen et al (2016) led Izzo et al (2018) to de-rive low values of reddening from the relations of van denBergh amp Younger (1987) The broadband photometry re-ported in Chen et al (2016) around maximum are different(by sim 08mag) from simultaneous SMARTS measurements(see Table C1)

For the purpose of our analysis we will use the valueof AV = 067 plusmn 002 derived from the fitting of the LETGChandra spectra which is taken at a late epoch (day 340)and is most sensitive to the column density In some caseswe make use of both extreme values (AV = 067 and 289)derived from the X-ray and optical spectroscopy

512 Distance and eruption amplitude

The so-called ldquoMaximum Magnitude-Rate of Declinerdquo(MMRD) relations (see eg McLaughlin 1939 Livio 1992Della Valle amp Livio 1998 Downes amp Duerbeck 2000) wereconsidered useful for deriving distances to novae These arebased on the assumption that the mass of the WD (MWD) isthe only factor that controls the brightness of the eruptionwhich led to the use of novae as standard candles The rela-tions have been questioned theoretically (see eg FerrareseCote amp Jordan 2003 Yaron et al 2005) and subsequentlyproved unreliable following the discovery of faint-fast novaein the M31 (Kasliwal et al 2011) and M87 (Shara et al2016 2017a) It is now accepted that several factors influ-ence the eruption of a nova including the accretion rate themass of the accreted envelope the chemical composition andthe temperature of the WD in addition to MWD Buscombeamp de Vaucouleurs (1955) have suggested that all novae de-cline to the same magnitude around 15 days after maximumlight This method is also is also considered uncertain espe-cially for fast novae (see eg Darnley et al 2006 Shara et al2017b and references therein) Using the MV15 method weobtain a distance estimate d ≃ 34+13

minus10 kpc with AV = 289

plusmn 025 and d ≃ 94+23minus18 kpc with AV = 067 plusmn 002 We do

not adopt any of the distances derived using the methodsdiscussed above due to their great uncertainty

The nova is included in Gaia DR1 (data between 2014July 25 and 2015 September 16) at G = 1895 and in DR2(data between 2014 July 25 and 2016 May 23) at G = 2066DR2 also provides a parallax of 127plusmn 151 which does notcontribute anything to our discussion of the distance

For the purpose of this discussion we assume a distanceof 10 kpc which should probably be regarded as an upperlimit Distance dependent parameters (eg radius and lu-minosity see Appendix A) can be easily adjusted for othervalues Future Gaia data releases will probably improve theparallax and its associated uncertainty

Data collection for DR2 stopped just before the novaeruption The change of ∆G sim 17 between DR1 and DR2suggests the nova may have faded before the eruption Atmaximum nova V407 Lup was at V 60 Therefore theamplitude of the eruption is sim 15mag Note that the ampli-tude of the eruption is determined by both the energetics ofthe eruption itself and by the luminosity of the donor duringquiescence Such a large amplitude is expected for novae ofthe same speed class as V407 Lup with a main-sequencecompanion (see eg fig 54 in Warner 1995) Large ampli-tude eruptions have been observed in a few other novae in-cluding V1500 Cyg CP Pup and GQ Mus (see eg Warner1985 Diaz amp Steiner 1989)

52 Colour-magnitude evolution

In Fig 29 we present a colour-magnitude diagram (CMD)illustrating the evolution of MV as a function of (B minus V )0Such diagrams have been used to constrain the nature ofthe secondary star in CNe Darnley et al (2012) proposed anew classification system for CNe based exclusively on theevolutionary state of the secondary star a main-sequencestar (MS-Nova) a sub-giant star (SG-Nova) or a red gi-ant star (RG-Nova) Hachisu amp Kato (2016a) have shownthat there is a difference between the evolutionary track onthe CMD of MS- and SG-novae compared to that of RG-novae For the latter the track follows a vertical trend be-tween (B minus V )0 = minus003 and (B minus V )0 = 013 The firstvalue represents the intrinsic colour of optically thin free-freeemission (free-free emission from an ldquooptically thin windrdquo ndashwinds that are accelerated outside the photosphere) whilethe second value represents the intrinsic colour of opticallythick free-free emission (free-free emission from an ldquoopti-cally thick windrdquo ndash winds that are accelerated deep insidethe photosphere see fig 5 in Hachisu amp Kato 2014) Novaewith main-sequence or sub-giant companions show a trackthat evolves blue-ward initially going from maximum lightto reach a turning point around the start of the nebularphase After the turning point the track evolves red-wardin the opposite direction (see fig 7 in Darnley et al 2016)

Fig 29 demonstrates that the evolution of the colourson the CMD follows a similar track to that shown by MS-and SG-novae Initially the colours evolve blue-ward be-tween day 2 and day 13 when the broadband observationsstopped for more than a hundred days (Sun constrained)and only resumed at day 117 when the colours had al-ready started evolving red-ward Hence the turning-point

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 25: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 23

Figure 29 Colour-magnitude diagram of V407 Lup showingMV against (B minus V )0 (red dots) at a distance d = 10 kpc and

E(BminusV ) = 022(AV = 067) The evolution of MV against timeis going vertically from top to bottom (as indicated by the blackarrow) and it spans from day sim 2 to day sim 13 post-eruption andthen from day sim 117 to day sim 283 Between these two intervals aturning-point took place in the evolutionary track of the coloursswitching from a blue-ward track to a red-ward track Aroundthis turning point the nova must have entered the nebular phase

and the SSS phase must have started as indicated in the dia-gram The blue horizontal dashed line indicates the estimated

maximum absolute magnitude The blue vertical line indicatesthe intrinsic colours for an optically thick wind free-free emission

(see eg Hachisu amp Kato 2014 Darnley et al 2016 and referencestherein)

and eventually the start of the nebular phase and the SSSphase have occurred between day 13 and day 117

Although we lack broadband observations of the sourceduring quiescence the evolution of the colour-magnitudetrack suggests that the companion star is likely to be a main-sequence or a sub-giant star This is consistent with (1) thequiescent magnitude measured by Gaia (G sim 19) At thismagnitude and with a giant secondary the distance to thenova would be unreasonable (d sim 30 ndash 40 kpc) (2) the shortorbital period (357 h) of the binary system Such a shortorbital period rules out a giant or sub-giant secondary

It is worth noting that the first two data points on theCMD measured at sim 05 - 15 d after maximum representan intrinsic colour (BminusV )0 sim 06 This is inconsistent withthe intrinsic colours of novae around maximum and is anindication that the adopted value of AV = 067 might betoo low a value (see Section 511)

53 Mass of the WD and ejected envelope

The behaviour of the nova eruption is governed by severalfactors such as the WD mass temperature and chemicalcomposition as well as the chemical composition of the ac-creted matter (Yaron et al 2005 Hillman et al 2014 Sharaet al 2017a) However the mass of the WD and conse-

quently the mass and velocity of the ejected envelope areconsidered the main factors that control the decline rate ofthe light-curve The more massive the WD is the less ac-creted material is required to trigger a TNR and hence themass of the ejected envelope is small and the light-curve de-cline time is short The opposite is equally true With t2 6

29 days the mass of the WD of V407 Lup is expected to beamp 125M⊙ (Hillman et al 2016)

Shara et al (2017a) derived an empirical relation be-tween the mass of the accreted envelope and the decline timet2 using nova samples from M87 M31 the large Magellaniccloud and the Galaxy logMenv = 0825 log(t2) minus 618 Us-ing this relation we derive Menv sim 16 times 10minus6 M⊙ typicalfor very fast novae (Hachisu amp Kato 2016b)

Sala amp Hernanz (2005) derived relations between theSSS temperature and the mass of the WD For kT sim80plusmn10 eV (see Fig 25) and for Menv sim 16times10minus6 M⊙ we es-timate a MWD sim 11 ndash 13M⊙ (see fig 1 in Sala amp Hernanz(2005)) We also estimate a value of MWD sim 12 ndash 13M⊙

based on fig 1 in Wolf et al (2014) Note that the tempera-ture values given in Fig 25 are model dependent thereforecaution is needed when interpreting the aforementioned es-timates

The SSS turn-on time is also a useful indication for themass of the ejected envelope However it is not possible toconstrain when the SSS emission emerged as the Swift mon-itoring was interrupted due to solar constraints and onlyresumed after the SSS has already turned on On the otherhand the SSS turn-off time (toff) which is strongly corre-lated with the WD mass is better constrained by the X-rayobservations (toff sim 300 d) This time is inversely propor-tional to the WD mass as toff prop Mminus63

WD (MacDonald 1996)After an X-ray survey in M31 Henze et al (2014) derivedempirical relations between the SSS parameters Their rela-tions suggest that very fast novae such as V407 Lup (witht2 lt 5 d) are expected to have ton 25 d and toff 100 dThis is much shorter than the observed toff for V407 Lup(see Section 54 for further discussion)

Based on the optical light-curve decline rate and thetemperature of the SSS we estimate a WD mass between11 and 13M⊙

54 Resumption of accretion

Although the XMM-Newton X-ray spectrum (taken on day168) is originating from H burning on the WD surface thespin-modulated signal observed in the XMM-Newton X-raydata is possibly indirect evidence for accretion restorationonto the polar regions of the WD as early as 168 days post-eruption This accretion is either igniting H burning at thepoles or resulting in an inhomogeneous atmosphere leadingto the observed modulation

Typically the spin-modulated signals seen in the X-raylight-curves of IPs are due to the obscuration caused bythe accretion curtains and possibly by the changing pro-jected area of the hard X-ray accretion-shock-regions nearthe pole of the WD However the spin-modulated XMM-

Newton light-curve of V407 Lup is observed during the SSSwhere the energy output is likely to be much higher thanwhat could result from accretion (ie the release of gravita-tional potential energy by accretion on to the WD gives amuch lower luminosity per unit mass accreted than nuclear

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 26: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

24 Aydi et al

burning does) Hence there is a possibility that the spin-modulated X-ray emission seen in the XMM-Newton RGSdata is not simply due to the supersonic accretion shock(or absorption by an accretion curtain) however it is orig-inating from the burning of the accreting material as it ar-rives at the WD in magnetically confined regions close tothe magnetic poles possibly making the atmosphere hotternear the poles This might be responsible for the observedsmall modulation fraction of plusmn5 in the RGS data (Fig 7)while simultaneously the post-nova residual H burning onthe whole surface of the WD is likely responsible for themajority of the SSS flux This idea was examined by KingOsborne amp Schenker (2002) for a SSS in M31 as IP-likemagnetically controlled accretion deposits fuel near to themagnetic poles of the WD which burns immediately so giv-ing rise to local hotluminous regions on the WD surfacepossibly contributing to the modulation seen in the opticaland X-rays

Such a conclusion is well supported by the luminosityderived from the different models in Section 44 These lumi-nosities exceed the accretion luminosity by orders of mag-nitude (the accretion X-ray luminosity of an IP does notexceed a few 1033 erg sminus1 eg Bernardini et al 2012) Thismeans that most of the luminosity cannot originate onlyfrom accretion onto a restricted area on the WD surfaceTherefore the majority of the SSS luminosity is most likelycoming from the whole WD surface while a small fraction iscoming from a hotluminous region near the poles leadingto the observed modulation However this does not rule outthe possibility that absorption from an accretion curtain canalso contribute to the observed modulation

Deriving an accurate value for the fraction of the WDsurface which is responsible for the modulation is not achiev-able with the available data This value is dependent onconstraining the inclination of the system relative to theobserver the temperature ratio between the whole WDsurface and the hotluminous region and the phenomenaresponsible for the modulation which are all still uncer-tainunknown Eclipses in the light-curves are usually usedto derive the dimension of the fractional area emitting inthe X-ray and causing the modulation Such eclipses areabsent in the light-curves of V407 Lup An alternative ap-proach is to use the X-ray luminosity derived during the SSSphase noting the above caveat Fig 7 shows that the ampli-tude of the modulation is sim 10 therefore we assume that10 of the SSS luminosity is emitted by the hotluminousregion (due to burning newly accreted material near thepoles) In Section 44 we derive an X-ray luminosity of sim11times1038 erg sminus1 and a Teff sim 80 plusmn 10 eV Thus the emit-ting area is A sim L(10)σT

4eff sim 26plusmn 10times 1016 cm2 Hence

using the effective radius derived from the 1 component T-MAP model (Fig 27) the fractional area responsible for themodulation would be f sim 17plusmn06times 10minus2 These values arewithin the normal expectation of 0001 f 002 for anIP (eg Rosen 1992 James et al 2002) However using theeffective radius derived from the blackbody model (Fig 27)implies a fractional area of only f sim 3plusmn 12times 10minus4

The Pspin modulation seen in the Chandra data at day340 might also be due to localized H burning near the polesof freshly accreted material while the residual H burning onthe rest of the WD surface is fading This is supported bythe little variation in the hardness ratio and the modulation

of the harder X-rays (Fig 8) which suggest that the Pspin

modulation might not be due to typical absorption by an ac-cretion curtain Lanz et al (2005) first noticed that the WDmodel atmosphere that best fitted the spectrum of the non-nova CAL 83 supersoft source indicated a luminosity almosta factor of 10 higher than observed and they hypothesizedthat perhaps the emission originated in hot polar caps In astudy of the decline and return to quiescence of novae thatare suspected IPs in which the disk is disrupted and a mod-erately strong (few 106 Gauss) magnetic field channels theaccretion stream to the WD poles Zemko Mukai amp Orio(2015) and Zemko et al (2016) found that such novae seemto decline in flux while remaining SSS as if the emissionregion shrinks before any final cooling These authors haveattributed the supersoft-emission to H burning at heatedregions near the poles where accreting material impingesNova V4743 Sgr in eruption also showed supersoft X-rayflux modulated with the WD Pspin (see Dobrotka amp Ness2017 and numerous references therein)

Signatures of accretion have often been found in bothoptical and X-rays observations of post-eruption novae (egLeibowitz et al 1992 Retter Leibowitz amp Kovo-Kariti 1998Beardmore et al 2012 Ness et al 2012 Orio et al 2013) andit has also been hypothesized that the disk may not be fullydisrupted in the eruption It is also possible that nuclearburning is prolonged by renewed accretion and the super-soft X-ray source lasts for longer because it is refueled byirradiation induced mass transfer (Greiner Orio amp Schartel2003)Therefore we assume that during the SSS phase ofV407 Lup the strongly irradiated secondary is heated by theX-ray emission leading to an expansion of its outer layersThis will eventually enhance the mass transfer and increasethe accretion rate (Osborne 2015) The material can flowfrom the L1 Lagrangian point via a ballistic accretion streamto (possibly) a reformed truncated accretion disk Then itfollows the magnetic field lines and finally impacts the WDsurface near the magnetic poles possibly feeding the SSS(see Darnley et al 2017 Henze et al 2018 and referencestherein for a discussion about the surface hydrogen burn-ing feeding from a re-established mass accretion in the caseof the non-magnetic system M31N 2008-12a) In the caseof V407 Lup this might also explain the long-lasting SSSemission for what is likely a massive WD (see Section 53)

An X-ray power spectrum with a dominant signal atthe frequency of the WD Pspin such as the Chandra powerspectrum (Fig 6) taken on day 340 might be an indicationof a disk-fed accretion (Ferrario amp Wickramasinghe 1999)Norton et al (1992b) point out that in this case the spinmodulation dominates and the orbital modulation vanishesbecause ldquopassage through a disk washes out all dependenceon the orbital phase from the accretion processrdquo Thereforeit is very likely that a truncated accretion disk has reformedby the time of the Chandra observation or even before Nev-ertheless a clear indication of a disk-fed accretion would bevery-low amplitude variations of the spectral line radial ve-locities modulated over the spin frequency (Norton Beard-more amp Taylor 1996 Ferrario amp Wickramasinghe 1999 seeSection 414 for a discussion of the spectral line velocities)It is worth noting that an accretion disk might reform over afew orbital cycles (private communication with B Warner)

This is consistent with the flux derived from the X-ray grating data at day 340 (see Section 45) As mentioned

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 27: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 25

previously in this section there is a possibility that a portionof the WD burning surface is hidden This might be due toa reformed accretion disk and an accretion curtain aroundday 340 or due to localized H burning near the poles asdescribed previously (or both)

The emergence of narrow and moving He ii lines around250 days post-eruption is another indication that accretionhas resumed as these lines are possibly originating from anirradiated accretion streamcurtain (see eg Warner 1995Potter et al 2004 Zemko et al 2016) or from a reformedaccretion disk (see eg Mason et al 2012 Mason amp Walter2014 Munari Mason amp Valisa 2014) In addition at daysim 304 Balmer emission components with a FWHMsim 500km sminus1 emerge on top of the broad nebular lines showingchanges in radial velocity (see Section 4131) These linesmight also originate from a reformed accretion disk Fewclassical or recurrent novae have shown such narrow movingHe ii and Balmer lines after the eruption when the ejectabecome optically thin Walter amp Battisti (2011) suggestedthat this can be a sign of either the survival of the accretiondisk or accretion restoration soon after the eruption

We conclude that there is a possibility of accretionrestoration around 168 days post-eruption causing local-ized H burning andor temperature inhomogeneity on theWD surface and leading to the observed spin-modulated sig-nal The narrow lines observed in the optical spectra theirlow velocity amplitude and the opticalX-ray power spectradominated by the Pspin frequency point towards resumptionof accretion and the possibility of a reformed accretion diskAlthough it is possible that the accretion has resumed ear-lier than 168 days after eruption in V407 Lup it is not pos-sible to constrain when exactly that happened due to thelack of observations between days sim 17 and sim 150 (solarconstraints) It is also worth noting that features of accre-tion resumption might only be observable when the ejectabecome optically thin

We showed in Section 34 that the AAVSO optical andXMM-Newton X-ray light-curves are out of phase whenfolded over the Pspin However caution is required wheninterpreting this comparison particularly because (1) theerror (05 s) on the WD Pspin derived from the RGS data istoo large for the ephemeris to be extrapolated to the dateof the AAVSO observations (gt 150 days later) and for thephase to be known during these observations (2) the XMM-

Newton X-ray light-curve was taken during the SSS phasewhile the AAVSO data were taken when the surface H burn-ing had started fading considerably

55 The Intermediate Polar scenario

Several novae have been suggested to occur in IP systemsincluding V533 Her GK Per DD Cir V1425 Aql V4743Sgr and Nova Scorpii 1437 AD (see eg Warner amp Woudt2002 Bianchini et al 2003 Woudt amp Warner 2003 2004Zemko et al 2016 Potter amp Buckley 2018 and referencestherein) These systems either show spin-modulated opticaland X-ray light-curves strong He ii emission lines or spin-modulated circular polarization

There is a striking similarity between the He ii lines ofV407 Lup and the complex He ii lines seen in mCVs (RosenMason amp Cordova 1987 Schwope Mantel amp Horne 1997)Zemko et al (2016) have found multiple components blue

and red-shifted with the same velocity in the spectra ofV4743 Sgr which also have been found in GK Per an oldnova and IP and were attributed to emission from the ac-cretion curtains in the magnetosphere of the WD (Bianchiniet al 2003)

The strong He ii 4686 A line is usually seen in the spec-tra of IPs although this line can also be prominent in otherCV systems that host a very hot WD We measure the EWof the narrow component in Hβ in the last three HRS spec-tra when the contribution from the broad emission (fromthe ejecta) is minimal The ratio of EW(He ii 4686 A) andEW(Hβ) is sim 2 Such a value is in agreement with the ratioseen in magnetic CVs and IPs which is usually gt 04 (Silber1992)

Belle et al (2003) detected very narrow lines of Ov andNv in the UV spectra of the IP EX Hydra These high ion-ization very narrow (FWHM sim 60 km sminus1) lines are verysimilar to the ldquovery narrow linesrdquo that we detect in the op-tical spectra of V407 Lup The authors of this study sug-gest that such high ionization lines should originate froma very high temperature region which increases the proba-bility that the lines are not contaminated by emission fromthe accretion disk Therefore due to their small widths theysuggested that the lines are formed near the surface of theWD

The main feature that supports the IP scenario in V407Lup is the detection of the two periodicities in the X-rayUV and optical light-curves These periodicities are verylikely to represent the Porb of the binary and the Pspin ofa highly magnetized WD and therefore are compelling evi-dence that V407 Lup is an IP

The detection of a hard X-ray component in the late (ampday 300) Swift X-ray data also favours the IP scenario Inaddition the luminosity of the hard X-ray component in theSwift X-ray data taken around the end of January 2018 is inthe range observed in IP systems (see Section 43) Howeverthis has still to be confirmed when a better distance estimateis available

All of these point towards the scenario of a nova occur-ring in an IP system and an evidence for accretion restora-tion

6 SUMMARY AND CONCLUSIONS

We have presented multiwavelength observations of novaV407 Lup which was discovered on HJD 24576555 (2016September 240 UT Stanek et al 2016) The optical UVand X-ray data of this nova have led us to the followingconclusions

1 V407 Lup is a very fast nova With t2 6 29 d it is oneof the fastest known examples This is also an indicationthat the eruption took place on a high mass WD (amp 125M⊙)

2 Based on the evolution of the colours after the erup-tion the system is very likely to contain a main-sequencecompanion

3 Timing analysis of the optical UV and X-ray light-curves shows two periodicities of 357 h and 565 s Theseare interpreted as the Porb of the binary and the Pspin

of the WD respectively suggesting that the system is anIP

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 28: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

26 Aydi et al

4 The presence of modulation at the Pspin of the WD inthe X-ray and optical data from 168 days post-eruptionsuggests that an accretion towards the magnetic poleshas resumed leading to localized burning near the polarregions of the WD surface

5 The high resolution optical spectra exhibit very narrowand moving lines as early as 164 days post-eruption Suchmoving high-excitation lines are most likely associatedwith very hot regions within the inner binary system

6 At day 250 relatively strong narrow and moving linesof He ii emerge in the optical spectra Their profiles arecomplex similar to those observed in mCVs 285 dayspost eruption moderately narrow and moving Balmeremission components also emerge All of these lines arepossibly originating from accretion regions (accretionstream disk and curtain) within the inner binary sys-tem

7 Based on the peak temperature of the SSS emission(sim80plusmn10 eV) and the time taken for the optical bright-ness to fade by 2 magnitudes (t2 6 29 d) we concludethat the mass of the WD is in the range of 11 ndash 13M⊙

8 We constrain the turn-off time of the SSS to be around300 days post-eruption later than expected for a massiveWD The optical and X-ray data show evidence of someform of accretion restoration while the SSS is still onThe accretion of fresh material onto the WD surfacemay have been responsible for extending the duration ofthe SSS

As the system is evolving back to quiescence further ob-servations are required to constrain the following (i) IPsmight show optical circular polarization originating fromcyclotron radiation (see eg Buckley et al 1995 1997Potter amp Buckley 2018) therefore polarimetric observa-tions are encouraged to further investigate the IP identityof V407 Lup (ii) orbital-period-resolved and spin-period-resolved spectroscopy are both needed to derive any pe-riodicity from the optical emission lines and possibly toreveal the morphology of the system via Doppler tomog-raphy (iii) Future Gaia data releases might provide anaccurate distance estimate allowing us to better constrainthe properties of the system such as the absolute maxi-mum magnitude luminosities and evolutionary state ofthe companion

ACKNOWLEDGMENTS

A part of this work is based on observations made with theSouthern African Large Telescope (SALT) under the Largescience Programme on transient 2016-2-LSP-001 EA DBPAW SM and BM gratefully acknowledge the receipt of re-search grants from the National Research Foundation (NRF)of South AfricaAK acknowledges the National Research Foundation ofSouth Africa and the Russian Science Foundation (projectno14-50-00043)MO had support from a NASA-Smithsonian grant for Di-rector Discretionary Time Chandra observationsAPB KLP NPMK and JPO acknowledge support from theUK Space AgencyJS acknowledges support from the Packard Foundation This

paper is partially based on observations obtained at theSouthern Astrophysical Research (SOAR) telescope whichis a joint project of the Ministerio da Ciencia TecnologiaInovacaos e Comunicacaoes (MCTIC) do Brasil the USNational Optical Astronomy Observatory (NOAO) the Uni-versity of North Carolina at Chapel Hill (UNC) and Michi-gan State University (MSU)MJD acknowledges the partial support from the UK Scienceamp Technology Facilities Council (STFC)SS gratefully acknowledges partial support from NASAgrants to ASUWe thank B Warner and R E Williams and L Izzo forprivate discussionWe acknowledge the use of observations from the AAVSOInternational Database and we thank the observers who ob-tained these observationsWe thank an anonymous referee for useful comments

REFERENCES

Arnaud K A 1996 in Astronomical Society of the PacificConference Series Vol 101 Astronomical Data AnalysisSoftware and Systems V Jacoby G H Barnes J edsp 17

Aschenbach B 2002 in Proc SPIE Vol 4496 X-Ray Op-tics for Astronomy Telescopes Multilayers Spectrome-ters and Missions Gorenstein P Hoover R B eds pp8ndash22

Aydi E et al 2018 MNRAS 474 2679Ballester P 1992 in European Southern Observatory Con-ference and Workshop Proceedings Vol 41 EuropeanSouthern Observatory Conference and Workshop Pro-ceedings Grosboslashl P J de Ruijsscher R C E eds p177

Barnes S I et al 2008 in Proc SPIE Vol 7014 Ground-based and Airborne Instrumentation for Astronomy II p70140K

Beardmore A P et al 2012 AampA 545 A116Beardmore A P Page K L Osborne J P Orio M 2017The Astronomerrsquos Telegram 10632

Belle K E Howell S B Sion E M Long K S SzkodyP 2003 ApJ 587 373

Bernardini F de Martino D Falanga M Mukai K MattG Bonnet-Bidaud J-M Masetti N Mouchet M 2012AampA 542 A22

Bianchini A Canterna R Desidera S Garcia C 2003PASP 115 474

Bode M F et al 2016 ApJ 818 145Bode M F Evans A 2008 Classical NovaeBode M F Kahn F D 1985 MNRAS 217 205Bramall D G et al 2012 in Proc SPIE Vol 8446Ground-based and Airborne Instrumentation for Astron-omy IV p 84460A

Bramall D G et al 2010 in Proc SPIE Vol 7735Ground-based and Airborne Instrumentation for Astron-omy III p 77354F

Brinkman A C et al 2000a ApJ 530 L111Brinkman B C et al 2000b in Proc SPIE Vol 4012X-Ray Optics Instruments and Missions III TruemperJ E Aschenbach B eds pp 81ndash90

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 29: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 27

Buckley D A H Haberl F Motch C Pollard KSchwarzenberg-Czerny A Sekiguchi K 1997 MNRAS287 117

Buckley D A H Sekiguchi K Motch C OrsquoDonoghueD Chen A-L Schwarzenberg-Czerny A Pietsch WHarrop-Allin M K 1995 MNRAS 275 1028

Burrows D N et al 2005 Space Sci Rev 120 165

Buscombe W de Vaucouleurs G 1955 The Observatory75 170

Cao Y et al 2012 ApJ 752 133

Cash W 1979 ApJ 228 939

Chen P et al 2016 The Astronomerrsquos Telegram 9550

Clemens J C Crain J A Anderson R 2004 inProc SPIE Vol 5492 Ground-based Instrumentation forAstronomy Moorwood A F M Iye M eds pp 331ndash340

Crause L A et al 2014 in Proc SPIE Vol 9147 Ground-based and Airborne Instrumentation for Astronomy V p91476T

Crawford S M et al 2010 in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series Vol7737 Society of Photo-Optical Instrumentation Engineers(SPIE) Conference Series p 25

Darnley M J et al 2006 MNRAS 369 257

Darnley M J et al 2016 ApJ 833 149

Darnley M J et al 2017 ApJ 849 96

Darnley M J Ribeiro V A R M Bode M F HounsellR A Williams R P 2012 ApJ 746 61

Della Valle M Livio M 1998 ApJ 506 818

Della Valle M Pasquini L Daou D Williams R E 2002AampA 390 155

den Herder J W et al 2001 AampA 365 L7

Diaz M P Steiner J E 1989 ApJ 339 L41

Dobrotka A Ness J-U 2010 MNRAS 405 2668

Dobrotka A Ness J-U 2017 MNRAS 467 4865

Downes R A Duerbeck H W 2000 AJ 120 2007

Ferrarese L Cote P Jordan A 2003 ApJ 599 1302

Ferrario L Wickramasinghe D T 1999 MNRAS 309517

Ferrario L Wickramasinghe D T King A R 1993 MN-RAS 260 149

Gehrels N et al 2004 ApJ 611 1005

Greiner J Orio M Schartel N 2003 AampA 405 703

Hachisu I Kato M 2006 ApJS 167 59

Hachisu I Kato M 2014 ApJ 785 97

Hachisu I Kato M 2016a ApJ 816 26

Hachisu I Kato M 2016b ApJS 223 21

Hatzidimitriou D Reig P Manousakis A Pietsch WBurwitz V Papamastorakis I 2007 AampA 464 1075

Hellier C 2001 Cataclysmic Variable Stars

Henze M et al 2018 ArXiv e-prints

Henze M et al 2014 AampA 563 A2

Hillman Y Prialnik D Kovetz A Shara M M 2016ApJ 819 168

Hillman Y Prialnik D Kovetz A Shara M M NeillJ D 2014 MNRAS 437 1962

Hounsell R et al 2010 ApJ 724 480

Izzo L et al 2016 The Astronomerrsquos Telegram 9587

Izzo L et al 2018 MNRAS

James C H Ramsay G Mark Cropper M CBranduardi-Raymont G 2002 MNRAS 336 550

Jansen F et al 2001 AampA 365 L1

Jose J Shore S N 2008 in Classical Novae Bode M FEvans A eds pp 121ndash140

Kaluzny J Semeniuk I 1987 Acta Astron 37 349Kasliwal M M et al 2011 ApJ 730 134King A R Osborne J P Schenker K 2002 MNRAS329 L43

Kniazev A Y Gvaramadze V V Berdnikov L N 2016MNRAS 459 3068

Kotze E J Potter S B McBride V A 2016 AampA 595A47

Krautter J 2008 in Classical Novae Bode M F EvansA eds pp 232ndash251

Kuin N P M et al 2015 MNRAS 449 2514Kuin P 2014 UVOTPY Swift UVOT grism data reduc-tion Astrophysics Source Code Library

Lanz T Telis G A Audard M Paerels F RasmussenA P Hubeny I 2005 ApJ 619 517

Leibowitz E Orio M Gonzalez-Riestra R Lipkin YNess J-U Starrfield S Still M Tepedelenlioglu E 2006MNRAS 371 424

Leibowitz E M Mendelson H Mashal E Prialnik DSeitter W C 1992 ApJ 385 L49

Li K-L et al 2017 Nature Astronomy 1 697Livio M 1992 ApJ 393 516MacDonald J 1996 in Astrophysics and Space ScienceLibrary Vol 208 IAU Colloq 158 Cataclysmic Variablesand Related Objects Evans A Wood J H eds p 281

Mason E Ederoclite A Williams R E Della Valle MSetiawan J 2012 AampA 544 A149

Mason E Munari U 2014 AampA 569 A84Mason E Shore S N De Gennaro Aquino I Izzo L PageK Schwarz G J 2018 ApJ 853 27

Mason E Walter F M 2014 in Astronomical Society ofthe Pacific Conference Series Vol 490 Stellar Novae Pastand Future Decades Woudt P A Ribeiro V A R Meds p 199

Mason K O et al 2001 AampA 365 L36Matthews J H Knigge C Long K S Sim S A Higgin-bottom N 2015 MNRAS 450 3331

McLaughlin D B 1939 Popular Astronomy 47 410Munari U Mason E Valisa P 2014 AampA 564 A76Munari U Ribeiro V A R M Bode M F Saguner T2011 MNRAS 410 525

Munari U Zwitter T 1997 AampA 318 269Murray S S et al 1997 in Proc SPIE Vol 3114 EUVX-Ray and Gamma-Ray Instrumentation for AstronomyVIII Siegmund O H Gummin M A eds pp 11ndash25

Ness J-U et al 2009 AJ 137 4160Ness J-U et al 2012 ApJ 745 43Ness J-U et al 2017 The Astronomerrsquos Telegram 10756Norton A J 1993 MNRAS 265 316Norton A J Beardmore A P Taylor P 1996 MNRAS280 937

Norton A J McHardy I M Lehto H J Watson M G1992a MNRAS 258 697

Norton A J Watson M G 1989 MNRAS 237 853Norton A J Watson M G King A R Lehto H JMcHardy I M 1992b MNRAS 254 705

Norton A J Wynn G A Somerscales R V 2004 ApJ614 349

Orio M et al 2013 MNRAS 429 1342Osborne J P 1988 Mem Soc Astron Italiana 59 117

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 30: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

28 Aydi et al

Osborne J P 2015 Journal of High Energy Astrophysics7 117

Osborne J P et al 2001 AampA 378 800Osborne J P et al 2011 ApJ 727 124Page K L Osborne J P Wagner R M Beardmore A PShore S N Starrfield S Woodward C E 2013 ApJ768 L26

Patterson J 1994 PASP 106 209Payne-Gaposchkin C 1964 The galactic novaePietsch W Henze M Haberl F Hernanz M Sala GHartmann D H Della Valle M 2011 AampA 531 A22

Potter S B Buckley D A H 2018 MNRAS 473 4692Potter S B Romero-Colmenero E Watson C A BuckleyD A H Phillips A 2004 MNRAS 348 316

Poznanski D Prochaska J X Bloom J S 2012 MNRAS426 1465

Pretorius M L Mukai K 2014 MNRAS 442 2580Prieto J L 2016 The Astronomerrsquos Telegram 9564Rauch T Orio M Gonzales-Riestra R Nelson T StillM Werner K Wilms J 2010 ApJ 717 363

Retter A Leibowitz E M Kovo-Kariti O 1998 MNRAS293 145

Roming P W A et al 2005 Space Sci Rev 120 95Rosen S R 1992 MNRAS 254 493Rosen S R Mason K O Cordova F A 1987 MNRAS224 987

Sala G Hernanz M 2005 AampA 439 1061Schaefer B E 2010 ApJS 187 275Schlafly E F Finkbeiner D P 2011 ApJ 737 103Schwarz G J et al 2011 ApJS 197 31Schwarz G J et al 2015 AJ 149 95Schwope A D Mantel K-H Horne K 1997 AampA 319894

Shara M M et al 2017a ApJ 839 109Shara M M et al 2016 ApJS 227 1Shara M M et al 2017b ArXiv e-printsShields G A Ferland G J 1978 ApJ 225 950Shore S N De Gennaro Aquino I Schwarz G J Au-gusteijn T Cheung C C Walter F M Starrfield S2013 AampA 553 A123

Shore S N Kenyon S J Starrfield S Sonneborn G1996 ApJ 456 717

Shore S N et al 2016 AampA 590 A123Silber A D 1992 PhD thesis MASSACHUSETTS IN-STITUTE OF TECHNOLOGY

Smith R K Brickhouse N S Liedahl D A RaymondJ C 2001 ApJ 556 L91

Stahl O Kaufer A Tubbesing S 1999 in AstronomicalSociety of the Pacific Conference Series Vol 188 Opti-cal and Infrared Spectroscopy of Circumstellar MatterGuenther E Stecklum B Klose S eds p 331

Stanek K Z et al 2016 The Astronomerrsquos Telegram 9538Starrfield S 1989 in Classical Novae Bode M F EvansA eds pp 39ndash60

Starrfield S Iliadis C Hix W R 2008 in Classical NovaeBode M F Evans A eds pp 77ndash101

Starrfield S Iliadis C Hix W R 2016 PASP 128 051001Stockman H S Schmidt G D Lamb D Q 1988 ApJ332 282

Strope R J Schaefer B E Henden A A 2010 AJ 14034

Struder L et al 2001 AampA 365 L18

Talavera A 2009 ApampSS 320 177Tody D 1986 in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series Vol 627 Instrumen-tation in astronomy VI Crawford D L ed p 733

Turner M J L et al 2001 AampA 365 L27van den Bergh S Younger P F 1987 AampAS 70 125van Rossum D R 2012 ApJ 756 43Wagner R M Depoy D L 1996 ApJ 467 860Walter F M Battisti A 2011 in Bulletin of the AmericanAstronomical Society Vol 43 American Astronomical So-ciety Meeting Abstracts 217 p 33811

Walter F M Battisti A Towers S E Bond H EStringfellow G S 2012 PASP 124 1057

Warner B 1985 MNRAS 217 1PWarner B 1995 Cambridge Astrophysics Series 28Warner B Woudt P A 2002 PASP 114 1222Weisskopf M C Tananbaum H D Van Speybroeck L POrsquoDell S L 2000 in Proc SPIE Vol 4012 X-Ray Op-tics Instruments and Missions III Truemper J E As-chenbach B eds pp 2ndash16

Werner K Rauch T Kruk J W 2007 AampA 474 591Williams R E Hamuy M Phillips M M HeathcoteS R Wells L Navarrete M 1991 ApJ 376 721

Wilms J Allen A McCray R 2000 ApJ 542 914Wolf W M Bildsten L Brooks J Paxton B 2014 ApJ782 117

Woudt P A Warner B 2003 MNRAS 340 1011Woudt P A Warner B 2004 in Astronomical Society ofthe Pacific Conference Series Vol 315 IAU Colloq 190Magnetic Cataclysmic Variables Vrielmann S CropperM eds p 39

Yaron O Prialnik D Shara M M Kovetz A 2005 ApJ623 398

Young P J Corwin Jr H G Bryan J de VaucouleursG 1976 ApJ 209 882

Zemko P Mukai K Orio M 2015 ApJ 807 61Zemko P Orio M Luna G J M Mukai K Evans P ABianchini A 2017 MNRAS 469 476

Zemko P Orio M Mukai K Bianchini A Ciroi SCracco V 2016 MNRAS 460 2744

Zhu H Tian W Li A Zhang M 2017 MNRAS 4713494

APPENDIX A DISTANCE DEPENDENT

PARAMETERS

The luminosity and effective radius at a certain distance d(in kpc) can be converted using

L(d) = L(10kpc)times 001 d2

R(d) = R(10kpc)times 01 d

where L(10kpc) andR(10kpc) are the luminosity and effectiveradius derived assuming a distance d = 10 kpc

Table A1 presents the luminosities and effective radiiderived from the modelling of the XMM-Newton RGS spec-trum for different distance assumptions Table A2 presentsthe luminosities derived from the modelling of the Chandra

LETG spectrum for different distance assumptions

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 31: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 29

Table A1 The luminosities and effective radii derived from thefitting of the blackbody model (BB) 1 component atmospheremodel (1-T) and 2 components atmosphere model (2-T) to theXMM-Newton RGS spectrum for different distance assumptions

d (kpc) 3 5 7

L(BB) (erg sminus1) 24 times 1038 67 times 1038 13times 1039

Reff(BB) (km) 15 times 104 26 times 104 36 times 104

L(1-T) (erg sminus1) 29 times 1037 80 times 1037 15 times 1038

Reff(1-T) (km) 21 times 103 35 times 103 49 times 103

L(2-T) (erg sminus1) 99 times 1036 27 times 1037 54 times 1037

R1(2-T) (km) 18 times 103 29 times 103 41 times 103

R2(2-T) (km) 36 times 103 60 times 103 84 times 103

Table A2 The luminosities derived from the fitting of models(1) and (2) to the Chandra LETG spectrum for different distanceassumptions

d (kpc) 3 5 7

L(1) (erg sminus1) 51 times 1035 14 times 1036 28 times 1036

L(2) (erg sminus1) 70 times 1036 19 times 1037 38 times 1037

APPENDIX B COMPLEMENTARY PLOTS

In this Appendix we present complementary plots

APPENDIX C TABLES

In this Appendix we present the tables

APPENDIX D OBSERVATION LOG

In this Appendix we list the log of the observations TheSMARTS and the Swift UVOT photometry can be found onthe electronic version The SMARTS time series photometryis available on the SMARTS online atlas at 11

11 httpwwwastrosunysbedufwalterSMARTSNovaAtlas

nsmc2016nsmc2016html

Table C1 V and Vis measurements around maximum from dif-ferent detectors

HJD (tminus t0) Magnitude Uncertainty Band Source(days)

245765550 00 910 0010 V ATel 9538245765593 043 682 0075 V AAVSO245765624 074 643 0050 V ATel 9550245765657 102 632 mdash Vis AAVSO245765690 110 560 mdash Vis AAVSO245765722 172 633 0060 V ATel 9550245765745 195 650 mdash Vis AAVSO245765746 196 684 mdash Vis AAVSO245765748 197 680 mdash Vis AAVSO245765748 197 714 0001 V SMARTS

245765751 201 670 mdash Vis AAVSO245765789 239 662 0050 Vis AAVSO

245765792 242 670 mdash Vis AAVSO245765841 291 710 mdash Vis AAVSO

245765846 296 692 mdash Vis AAVSO245765848 298 690 mdash Vis AAVSO245765848 298 726 0001 V SMARTS245765899 349 720 mdash Vis AAVSO245765941 391 750 mdash Vis AAVSO245765942 392 750 mdash Vis AAVSO

245765944 394 760 mdash Vis AAVSO245765946 396 754 mdash Vis AAVSO245765946 396 740 mdash Vis AAVSO245765998 448 785 mdash Vis AAVSO245766023 473 770 mdash Vis AAVSO245766041 491 820 mdash Vis AAVSO245766042 492 800 mdash Vis AAVSO

245766043 493 810 mdash Vis AAVSO245766046 496 800 mdash Vis AAVSO

245766046 496 770 mdash Vis AAVSO245766048 498 788 0080 V ATel 9564

245766050 500 798 0001 V SMARTS245766092 542 870 mdash Vis AAVSO245766102 552 810 0012 V AAVSO245766120 570 830 mdash Vis AAVSO245766140 290 860 mdash Vis AAVSO245766141 591 850 mdash Vis AAVSO245766148 598 856 0001 V SMARTS

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 32: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

30 Aydi et al

Figure B1 The BVRI (top) and JHK (bottom) photometry from SMARTS as a function of days since eruption colour and symbol

coded as indicated in the legend We present a zoomed sub-plot to show in more detail the evolution of the light-curve during the earlydecline Day to day variability appear in the light-curve after t3 (marked by a red arrow)

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 33: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 31

Figure B2 A direct comparison between the SMARTS V -band (top) Swift UVOT uvw2 (middle) and Swift X-ray light-curves

(bottom) from day 100 The blue dashed lines represent the dates of the SALT HRS observations respectively The black dashed linerepresents the date of the XMM-Newton observations The red solid lines represent the median dates of the first four and second four

UV grism spectra The green dotted line represents the date of the Chandra observation

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 34: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

32 Aydi et al

Figure B3 Top A sample of the wind-type atmosphere models (van Rossum 2012) fitting to the XMM-Newton RGS spectrum TheTeff characterizing each model is indicated on the plot Bottom best fit wind-type atmosphere model to the Chandra LETG spectrumfor various wind asymptotic velocities vinfin and mass-loss rates M as indicated on the plot In all panels the models in blue are compared

to the observations in black

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 35: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 33

Table C2 Spectral line identification of the optical SALT and SOAR spectra We list the EW FWHM and integrated flux of the linesfor which an estimate was possible

Line λ0 EW (λ) FWHM Flux (tminus t0)(A) (km sminus1) 10minus13erg cmminus2sminus1

[Nev] 3346 minus55plusmn 5 2400plusmn100 30 plusmn02 269[Nev] 3426 minus160plusmn 10 2400plusmn100 80 plusmn05 269Ovi 3811 minus18plusmn 02 400plusmn 50 008 plusmn002 269

[Ne iii] 3869 minus75plusmn 10 2400plusmn100 27 plusmn02 269[Ne iii] 3968 minus35plusmn 5 2400plusmn100 12 plusmn02 269

Hδ 4102 minus10plusmn 2 2600plusmn100 034 plusmn005 269

[O iii] 4363 minus450plusmn 50 2400plusmn100 ndash 164Hγ 4341 ndash ndash ndash 164

He ii 4542 minus08plusmn 02 270plusmn 50 ndash 263He ii 4686 minus87plusmn 10 435 plusmn 50 ndash 250

He i 4713 ndash ndash ndash 164[Nevi] 4721 ndash ndash ndash 164

Hβ 4861 minus100plusmn20 2850plusmn100 ndash 164[O iii] 4959 ndash ndash ndash 164[O iii] 5007 ndash ndash ndash 164Ovi 5290 07plusmn 02 120plusmn30 ndash 164He ii 5412 minus38plusmn 05 460plusmn 50 ndash 250

[O i] 5577 ndash ndash ndash 164[Fevi] 5631 ndash ndash ndash 164

[Fevi] 5677 ndash ndash ndash 164[N ii] 5755 minus80plusmn 10 2800plusmn100 ndash 164He i 5876 minus20plusmn 5 2400plusmn100 ndash 164

[Fevii] 6087 minus35plusmn 5 3100plusmn100 ndash 164Ovi 6200 ndash ndash ndash 164[O i] 6300 minus30plusmn 5 ndash ndash 164[O i] 6364 minus25plusmn 5 ndash ndash 164Hα 6563 minus575 plusmn2 0 2900plusmn100 ndash 164He i 7065 minus25plusmn 5 2800plusmn100 ndash 164

[O ii] 732030 ndash ndash ndash 164Ne iv 7716 minus11plusmn 02 150 plusmn 30 ndash 164

[Fex] 7892 minus30plusmn 5 2800plusmn100 ndash 164He ii 8237 ndash ndash ndash 164

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 36: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

34 Aydi et al

Table C3 The FWHM of the broad Hα and Hβ In the lastthree spectra the narrow features dominate the broad nebularemission hence we exclude them from the fitting

HJD tminus t0 FWHM (Hα) FWHM (Hβ)(days) plusmn 100 (km sminus1)

245781949 164 2900 2850245786438 209 2800 2800245790552 250 2700 2600

245791847 263 2650 2400245793443 279 2650 2300

245794041 285 2200 2400

245795936 304 ndash ndash245796435 309 ndash ndash245797232 317 ndash ndash

Table C4 The radial velocity of the ldquomoderately narrowrdquo Hβand Hα features These measurements were done by cursor posi-tion when possible

HJD tminus t0 Vrad(Hβ) Vrad (Hα)(days) plusmn 50 (km sminus1)

245790552 250 minus210 minus230

245791847 263 minus215 minus235245793443 279 ndash minus220

245794041 285 minus20 minus100245795936 304 minus20 minus20

245796435 309 ndash ndash245797232 317 10 minus20

Table C5 The FWHM of [O iii] 4363 A

HJD tminus t0 FWHM(days) plusmn 100 (km sminus1)

245781949 164 2950245786438 209 2700

245790552 250 2400

245791847 263 2200245793443 279 2300245794041 285 2200245795936 304 2100245796435 309 1800245797232 317 1600

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 37: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 35

Table C6 The radial velocity (Vrad) FWHM and EW derived from a single component fitting of the He ii lines

HJD t = tminus t0 Phase Vrad(He ii 4686) Vrad(He ii 5412) Vrad(He ii 4542)

(days) plusmn 30 (km sminus1)

245790552 250 0088 minus210 minus205 minus240245791847 263 0072 minus210 minus177 minus210245792451 270 0583 minus175 ndash ndash245793443 279 0269 minus155 minus140 minus135245794041 285 0464 minus25 minus20 minus40245795936 304 0766 minus70 minus100 minus100245796435 309 0229 minus90 minus100 minus105245797232 317 0762 0 minus20 minus60245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

HJD tminus t0 Phase FWHM (He ii 4686) FWHM (He ii 5412) FWHM (He ii 4542)

(days) plusmn 50 (km sminus1)

245790552 250 0088 435 460 ndash245791847 263 0072 470 550 270245792451 270 0583 550 ndash ndash245793443 279 0269 470 530 260245794041 285 0464 400 420 360245795936 304 0766 510 420 380245796435 309 0229 430 500 380245797232 317 0762 410 420 380245813984 484 0033 570 ndash ndash245814085 485 0806 465 ndash ndash

HJD tminus t0 Phase EW (He ii 4686) EW (He ii 5412) EW (He ii 4542)

(days) plusmn 1 (A) plusmn 05 (A) plusmn 02 (A)

245790552 250 0088 minus87 minus38 minus08245791847 263 0072 minus88 minus42 minus08

245792451 270 0583 minus90 ndash ndash245793443 279 0269 minus115 minus30 minus11245794041 285 0464 minus190 minus46 minus15245795936 304 0766 minus180 minus53 minus20245796435 309 0229 minus141 minus50 minus19245797232 317 0762 minus96 minus30 minus16

245813984 484 0033 minus90 ndash ndash245814085 485 0806 minus95 ndash ndash

Table C7 The radial velocity (Vrad) and FWHM of the mediumwidth component (MWC) of the He ii 4686 A line

HJD tminus t0 Phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0142 ndash ndash245786438 209 0708 ndash ndash245790552 250 0043 -200 240245791847 263 0029 -215 260245793443 279 0234 -30 340245794041 285 0418 -125 230245795936 304 0740 -170 265245796435 309 0205 +15 415

245797232 317 0741 -165 300

Table C8 The radial velocity (Vrad) and FWHM of the smallwidth component (SWC) of the He ii 4686 A line

HJD tminus t0 Orbital phase Vrad FWHM(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 ndash ndash245790552 250 0088 -35 70245791847 263 0072 -5 65245793443 279 0269 -205 120245794041 285 0464 -25 ndash245795936 304 0766 -10 106245796435 309 0229 -165 196

245797232 317 0762 +40 200

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 38: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

36 Aydi et al

Table C9 The full base width of the broad base component(BBC) of the He ii 4686 A line

HJD tminus t0 Orbital phase [λindash λf ] ∆λ(days) plusmn 10 (A)

245781949 164 0151 ndash ndash

245786438 209 0753 ndash ndash245790552 250 0088 [4671ndash 4691] 20

245791847 263 0072 [4671ndash 4691] 20

245793443 279 0269 [4671ndash 4691] 20245794041 285 0464 [4677ndash 4697] 20245795936 304 0766 [4671ndash 4693] 22245796435 309 0229 [4670ndash 4692] 22245797232 317 0762 [4670ndash 4693] 23

Table C10 The radial velocity (Vrad) of the ldquomoderately nar-rowrdquo Ovi 5290 A and 6200 A lines

HJD tminus t0 Phase Vrad(5290) Vrad(6200)(days) plusmn 30 (km sminus1)

245781949 164 0151 ndash ndash245786438 209 0753 5 ndash245790552 250 0088 -260 -165245791847 263 0072 -265 ndash

245793443 279 0269 -115 -100

245794041 285 0464 -75 -65245795936 304 0766 -255 -185245796435 309 0229 -80 -80245797232 317 0762 ndash ndash

Table C11 The radial velocity FWHM and EW of the ldquoverynarrow linesrdquo Ne iv 44984 A (top) Ovi 5290 A (middle) andNe iv 77159 A (bottom) The last two measurements (days 304and 309) the lines merge with neighbouring ldquomoderately narrow

linesrdquo and therefore the FWHM increases considerably

HJD tminus t0 Phase Vrad FWHM EW(days) plusmn 30 (km sminus1) plusmn 02 (A)

245781949 164 0151 minus5 120 minus04245786438 209 0753 minus130 ndash ndash245790552 250 0088 15 90 minus04245791847 263 0072 10 ndash ndash245793443 279 0269 minus50 140 minus08245794041 285 0464 minus150 85 minus03245795936 304 0766 minus90 230 minus03

245796435 309 0229 minus50 265 minus07245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 120 minus07245786438 209 0753 minus125 ndash ndash

245790552 250 0088 25 110 minus09245791847 263 0072 25 110 minus07

245793443 279 0269 minus35 110 minus09245794041 285 0464 minus145 60 minus04245795936 304 0766 minus75 340 minus25245796435 309 0229 minus30 280 minus15245797232 317 0762 ndash ndash ndash

245781949 164 0151 20 150 minus11245786438 209 0753 minus120 150 minus10245790552 250 0088 20 105 minus10245791847 263 0072 20 110 minus09245793443 279 0269 minus35 110 minus15245794041 285 0464 minus145 75 minus11

245795936 304 0766 minus80 250 minus23245796435 309 0229 minus40 310 minus35

245797232 317 0762 ndash ndash ndash

Table C12 Spectral line identifications of the UV-optical grismspectra

spectrum line vacuum lab noteswavelength (a) ID wavelength

1759 N iii] 17501855 Al iii 185518632141 N ii] 21432332 [O iii] 23322518 He ii 2511 blend

2635 unid Nevi 2626 2801 Mg ii 2800

2978 Neviii 29773135 O iii 31333346 [Nev] 33463424 [Nev] 34263868 [Ne iii] 38693970 [Ne iii] 39684104 H i 4103 weak4363 [O iii] 43634523 N iii 4517 weak4714 unid broad feature

4866 H i 48635010 [O iii] 495950075017 blend

(a) after shifting the spectra to best match the [Nev] doublet

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 39: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 37

Table C13 UV line measurements for the UV-optical grism spectra The numbers between brackets are days after t0

10minus10 Flux (erg cmminus2 sminus1 Aminus1

)nominal line MJD MJD MJD MJD MJD MJD MJD MJDwavelength ID 57820 57822 57825 57826 57838 57854 57857 57866

[+165] [+167] [+170] [+171] [+183] [+199] [+202] [+211]

2800 Mg II 976 149 184 106 77 96 76 1123133 O III 121 99 70 81 53 54 32 83346 [Ne V] 325 289 361 340 236 161 140 1553426 [Ne V] 903 768 944 835 593 411 367 3603869 [Ne III] 279 289 279 313 197 196 163 17

3968 [Ne III] 138 126 110 146 115 106 83 124

4363 [O III] 327 281 318 285 206 220 184 190

FWZI (A)

2800 Mg II 709 902 113 552 566 703 667 873133 O III 900 773 570 575 508 514 40 803346 [Ne V] 646 739 670 774 649 659 543 583426 [Ne V] 798 911 861 861 721 783 719 813869 [Ne III] 813 952 743 917 721 859 786 813968 [Ne III] 890 777 561 806 820 734 618 984363 [O III] 1136 882 116 974 957 995 900 103

Table C14 The parameters of the two-components TMAP

model used to fit the RGS spectrum F(abs) and F(unabs) rep-resent the absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 183 times1021

Teff1 (K) 1016 times106

Teff2 (K) 7930 times105

Velocity blue-shift (km sminus1 588F(1abs) (erg cmminus2 sminus1) 618 times10minus10

F(1unabs) (erg cmminus2 sminus1) 404 times10minus9

F(2abs) (erg cmminus2 sminus1) 598 times10minus10

F(2unabs) (erg cmminus2 sminus1) 526 times10minus9

Total F(abs) (erg cmminus2 sminus1) 122 times10minus9

Total F(unabs) (erg cmminus2 sminus1) 930 times10minus9

Table C15 The parameters of the first model (Model 1) used to

fit the Chandra LETG spectrum Tbvapec represents the temper-ature of the collisionally ionized plasma around the nova F(abs)

and F(unabs) represent the absorbed and unabsorbed fluxes re-spectively

Parameter Value

NH (cmminus2) 14 times1021

Teff (K) 586 times105

Tbvapec (eV) 808Atmosphere F(abs) (erg cmminus2 sminus1) 113 times10minus11

Atmosphere F(unabs) (erg cmminus2 sminus1) 129 times10minus10

CIE F(abs) (erg cmminus2 sminus1) 161 times10minus11

CIE F(unabs) (erg cmminus2 sminus1) 346 times10minus10

Total F(abs) (erg cmminus2 sminus1) 274 times10minus11

Total F(unabs) (erg cmminus2 sminus1) 475 times10minus10

Table C16 The parameters of the second model (Model 2) used

to fit the Chandra LETG spectrum F(abs) and F(unabs) representthe absorbed and unabsorbed fluxes respectively

Parameter Value

NH (cmminus2) 187 times1021

Teff (K) 636 times105

T(atm) (eV) 39

Total F(abs) (erg cmminus2 sminus1) 275times10minus11

Total F(unabs) (erg cmminus2 sminus1) 694 times10minus9

Table D1 A sample of the SMARTS BVRIJHK photometryThe time series photometry is available on the electronic versionand on the SMARTS atlas

HJD (tminus t0) Band Magnitude Instrument(days) (mag)

245765748 198 J 5245 0003245765748 198 H 4695 0002245765748 198 K 4850 0003245765748 198 I 6909 0000245765748 198 B 7944 0000245765748 198 V 7145 0000245765748 198 R 6953 0000245765848 298 J 5288 0003

245765848 298 H 4789 0002

245765848 298 K 4629 0003245765848 298 I 6750 0000245765848 298 B 8027 0000245765848 298 V 7260 0000245765848 298 R 6160 0001245766050 300 J 6032 0005245766050 300 H 5409 0003245766050 300 K 5492 0004245766050 300 I 6793 0000245766050 300 B 8107 0000245766050 300 V 7985 0001245766050 300 R 6752 0001

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 40: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

38 Aydi et al

Table D2 The SALT HRS spectroscopy observation log

HJD tminus t0 Exp time (blue) Exp time (red) Mode(days) (s)

245781949 164 1800 1800 LR245786438 209 1800 1800 LR245790552 250 1800 1800 LR245791847 263 1800 1800 LR245793443 279 1800 1800 LR245794041 285 1800 1800 LR245795936 304 1800 1800 LR245796435 309 1800 1800 LR245797232 317 1800 1800 LR

Table D3 A sample of the Swift observation log providing theobservation-averaged XRT and UVOT uvw2 photometric mea-surements The rest of the Swift observation log is available onthe electronic version

ObsID tminus t0 half-bin width XRT rate error uvw2 ∆(uvw2)(days) (mag)

00034741001 2916 0024 lt 0003 - - -00034741002 8210 0005 lt 0014 - - -00034741003 9141 0005 lt 0011 - - -00034741004 10300 0097 lt 0018 - - -00034741005 11492 0035 lt 0010 - - -00034741006 12162 0035 lt 0014 - - -

00034741007 13595 0005 lt 0015 - - -00034741008 14392 0006 lt 0012 - - -

00034741009 15157 0035 lt 0013 - - -

00034741010 16383 0006 66 $ (+26minus 19)times 10minus3 - -00034741012 19096 0066 lt 0039 - - -

00034741013 150423 0371 4203 014 1348 00200034741014 151657 0328 4974 017 - -00034741016 152575 0002 3351 027 1347 00200034741017 153784 0005 4182 021 1361 002

00034741018 154381 0005 3569 020 1357 00200034741019 155512 0002 3281 028 1349 002

00034741020 156768 0005 3762 020 1367 002

00034741021 157025 0001 3786 051 1356 00200034741028 161355 0001 4630 040 1380 00200034741030 164670 0001 3688 037 - -00034741031 164678 0005 3943 020 - -00034975002 167261 0001 4072 046 - -00034975001 167270 0007 3251 017 1376 00300034741034 168628 0105 3767 010 1384 00200034975003 170054 0001 2965 035 1389 002

00034975004 170056 0000 2609 048 - -

00034975005 171048 0001 2473 031 - -

Grade 0 events only were considered before the start of the solar observing constraintbecause of concern about optical loading (see text for details)

$ Possible detection

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log
Page 41: LJMU Research Onlineresearchonline.ljmu.ac.uk/id/eprint/8965/1/1807.00706v1.pdf · in cataclysmic variable (CV) systems. These are interacting binaries consisting of a WD primary

nova V407 Lup 39

Table D4 UV grism observations with lenticular filters

datetime MJD (tminus t0) exposure filter notes(UT) (days) time (s)

2017-03-07T1608 5781967 16467 8925 UGRISM (a)2017-03-10T0615 5782226 16726 2011 UVM22017-03-10T0619 5782226 16726 11070 UGRISM (a)2017-03-10T0638 5782227 16727 1036 UVW22017-03-13T0116 5782505 17005 2405 UVW22017-03-13T0121 5782505 17005 1035 UGRISM (a)2017-03-14T0108 5782604 17104 2855 UVM22017-03-14T0113 5782605 17105 11700 UGRISM (a)2017-03-14T0133 5782606 17106 988 UVW2

2017-03-26T0157 5783808 18308 2022 UVM22017-03-26T0201 5783808 18308 8393 UGRISM (b)

2017-03-26T0215 5783809 18309 766 UVW2

2017-04-11T0959 5785441 19941 1814 UVM22017-04-11T1003 5785441 19941 9013 UGRISM (b)2017-04-11T1018 5785442 19942 732 UVW22017-04-14T0015 5785701 20201 961 UVW22017-04-14T0017 5785701 20201 9997 UGRISM (b)2017-04-14T0034 5785702 20202 920 UVW22017-04-23T0404 5786616 21116 2228 UVM22017-04-23T0408 5786617 21117 8895 UGRISM (b)

2017-04-23T0423 5786618 21118 854 UVW2

(a) no offset on detector roll=112deg First order contamination lt 2050A

Zeroth orders contaminate [1750-1865][1950-2185][2410-2540] A(b) offset around 56prime on the detector roll = 130deg

Table D5 The XMM-Newton observation log taken on day 168

Exposure Number Instrument Observing Mode Filter Start of exposure Stop of exposure

001 MOS1 Full Frame MEDIUM FILTER 2017-03-11T110217 2017-03-11T170657002 MOS2 Timing Uncompressed MEDIUM FILTER 2017-03-11T110257 2017-03-11T170243

003 pn Timing MEDIUM FILTER 2017-03-11T113725 2017-03-11T170717004 RGS1 Spectro + Q NOT APPLICABLE 2017-03-11T110147 2017-03-11T170812005 RGS2 Spectro + Q NOT APPLICABLE 2017-03-11T110152 2017-03-11T170812011 OM OM Science User Defined VISIBLE GRISM 2 2017-03-11T110729 2017-03-11T121235012 OM OM Sci User Def FAST UVW1 2017-03-11T121236 2017-03-11T133102013 OM OM Sci User Def FAST UVW1 2017-03-11T133103 2017-03-11T141609014 OM OM Sci User Def FAST UVW1 2017-03-11T141610 2017-03-11T153428

015 OM OM Sci User Def FAST UVW1 2017-03-11T153429 2017-03-11T165255

  • 1 Introduction
  • 2 Observations and data reduction
    • 21 SMARTS observations and data reduction
    • 22 SALT high-resolution Echelle spectroscopy
    • 23 SOAR medium-resolution spectroscopy
    • 24 Swift X-ray and UV observations
      • 241 Swift XRT observations
      • 242 Swift UVOT observations
        • 25 XMM-Newton X-ray and UV observations
        • 26 Chandra X-ray observations
          • 3 Photometric results and analysis
            • 31 Optical light-curve parameters
            • 32 The Swift UVOT light-curve
            • 33 The XMM-Newton and Chandra UV and X-ray light-curves
            • 34 Timing analysis
              • 4 Spectroscopic results and analysis
                • 41 Optical spectroscopy
                  • 411 Line identification
                  • 412 Spectral classification and evolution
                  • 413 Line profiles
                    • 4131 Balmer lines
                    • 4132 Forbidden oxygen lines
                    • 4133 The He II moderately narrow lines
                    • 4134 The O VI moderately narrow lines
                    • 4135 Very narrow lines
                      • 414 Radial velocities
                      • 415 Sodium D absorption lines
                        • 42 Swift UVOT spectroscopy
                        • 43 Swift X-ray spectral evolution
                        • 44 XMM-Newton high-resolution X-ray spectroscopy
                        • 45 Chandra high-resolution X-ray spectroscopy
                          • 5 Discussion
                            • 51 Reddening distance and eruption amplitude
                              • 511 Reddening
                              • 512 Distance and eruption amplitude
                                • 52 Colour-magnitude evolution
                                • 53 Mass of the WD and ejected envelope
                                • 54 Resumption of accretion
                                • 55 The Intermediate Polar scenario
                                  • 6 Summary and conclusions
                                  • A Distance dependent parameters
                                  • B Complementary plots
                                  • C Tables
                                  • D Observation log