The$Inner$Milky$Way$Structure$ with$VVV$rsaito/pdf/REJKUBA.pdf · Fig.2. J,H and Ks magnitudes...
Transcript of The$Inner$Milky$Way$Structure$ with$VVV$rsaito/pdf/REJKUBA.pdf · Fig.2. J,H and Ks magnitudes...
The Inner Milky Way Structure with VVV
Marina Rejkuba ESO, Germany
Overview
• Red Clump stars as distance indicators – Dependence on age and metallicity
• Tracing the bar with RC stars in VVV • Inner bar flaFening and the bar orientaGon • The structure behind the bar
• Based on work published in: Gonzalez, O., et al., 2011, A&A 534, A3 Gonzalez, O., et al., 2011, A&A LeF 534, L14
Horizontal Branch Stars: Observations and Theory 3
Fig. 1 V , B!V CMD for the Carina dSph galaxy, withthe positions of red clump stars, red HB stars, the RRLyrae “gap,” and the blue HB indicated. The red clumpis associated with the younger turno!, at V " 23 mag,B!V " 0.25 mag. The other components derive from theold turno! whose presence is indicated by the faint subgiantbranch at V " 23.25 mag, B!V " 0.55 mag. Adapted fromMonelli et al. (2003).
2002). They are physically related to, but should notbe confused with, the so-called “red HB stars,” whichare less massive, significantly fainter and usually fallalong the “horizontal” part of the HB, as clearly shownin Figure 1.
A common mistake is to confuse “red clump” and“RGB bump” stars, but they are totally di!erent typesof stars: the latter are bona-fide red giants with apartially degenerate helium core that burn hydrogenin a thin shell surrounding the core (Thomas 1967).This hydrogen-burning shell moves outward in mass asthe star evolves up the RGB, and eventually reaches achemical discontinuity left behind by the previous, max-imum inward penetration of the hydrogen-rich convec-tive envelope, leading to a sudden replenishment of theshell with fresh nuclear fuel and a momentary reversalof evolutionary path on the RGB (see Salaris, Cassisi,& Weiss 2002 for a review and additional references).Recent papers discussing several di!erent aspects of redclump stars include those by Girardi (1999) and Salaris,Percival, & Girardi (2003).
Fig. 2 V , B!V (left panel) and U , U !V (right panel)CMDs for NGC 6752. In the left panel, the positions ofHBA, HBB, and EHB stars (collectively forming the “bluetail”) are indicated, along with the positions of possible gapsalong the distribution. In the right panel, besides the bluetail, the positions of the Grundahl et al. (1999) and Momanyet al. (2002, 2004) “jumps” are also indicated, along withthe approximate position where “blue hook” stars would befound if they were present in the cluster. Adapted fromMomany et al. (2002).
2.2 The red HB
As just stated, red HB stars are the lower-mass ana-logues of the red clump stars. They are commonlypresent in metal-rich (e.g., Armandro! 1988; Ortolaniet al. 1995) as well as relatively young globular clus-ters (e.g., Stetson et al. 1989; Buonanno et al. 1993).However, red HB stars can also be found in metal-poorsystems with “normal” ages, where they are usually in-terpreted as either the progeny of red giants that lostvery little mass on their ascent of the RGB, or as formerblue HB stars or RR Lyrae variables that have evolvedto the right on the color-magnitude diagram, now ap-proaching the end of their evolution as HB stars andthe beginning of the asymptotic giant branch (AGB)phase. Another route towards producing (admittedlyfewer) red HB stars is through the evolution of bluestraggler stars (BSS; Renzini & Fusi Pecci 1988; FusiPecci et al. 1992). Fusi Pecci et al. suggest, in fact,that a red HB star “of BSS origin” should be presentfor every ! 6 BSS present in a given globular cluster.(Note that, depending on the amount of mass lost bythe BSS on the RGB phase, at least some of these mightbetter classify as red clump stars.)
Catelan 2009
V
B-‐V
be presented in a separate paper (Pietrzynski et al. 2003).Here we use photometry obtained with the ESO NTT ontwo nights for three fields in the LMC and two fields in theSMC to study the magnitudes of red clump stars in thesegalaxies. Basic information on the data is given in Table 1.Observations, reductions, and calibrations of theseMagellanic Cloud data were performed in the same manneras for the data obtained for the Fornax and Carina dwarfgalaxies. The accuracy of the derived zero points for the twoNTT observing nights was better than 0.03 mag. It is worth-while noting that the instrumental systems at both the NTTand the VLT are identical, which obviously helped us toobtain very homogenous data for all four galaxies discussedin this paper.
3. MEAN K- AND J-BAND REDCLUMP MAGNITUDES
The near-infrared K versus J!K color-magnitude dia-grams around the red clump region are shown in Figure 1for our four target galaxies. The relatively large number ofred clump stars found in our four fields in Fornax (see Table1) allowed us to derive their mean K-band magnitudes indi-vidually for each field in this galaxy. We did this by selectingthe stars with K-band magnitudes in the range from 18.3 to20.3 mag and J!K colors in the range from 0.3 to 0.85 mag
and deriving the histogram of their K-band magnitudes inbins of 0.06 mag for each field. Then function (1), whichconsists of a Gaussian component representing the redclump stars distribution and a second-order polynomialfunction approximating the stellar background, was fittedto the data following a procedure originally introduced byPaczynski & Stanek (1998):
n"K# $ a% b"K ! Kmax# % c"K ! Kmax#2
% NRC
!RC
!!!!!!2"
p exp !"K ! Kmax#2
2!2RC
" #
: "1#
The same procedure was applied for deriving the meanred clump star magnitudes in the J band. The results arepresented in Table 2. It can be appreciated that the fourindependent measurements of the mean J- and K-band redclump magnitudes obtained on the two di!erent observingnights are in very good agreement. We therefore decided tomerge the data in order to improve the statistics. The fits offunction (1) to the histograms of the K and J magnitudesof red clump stars from the combined data are displayedin Figure 2. The corresponding results of hKi =19.215 & 0.013 and hJi = 19.687 & 0.014 were finallyadopted as the mean K- and J-band apparent magnitudes ofred clump stars in the Fornax dwarf galaxy.
Fig. 1.—K, J!K near-infrared color-magnitude diagrams for the LMC, SMC, Carina, and Fornax. Data have been combined from the di!erent fieldsobserved in each galaxy (see Table 1).
2496 PIETRZYNSKI, GIEREN, & UDALSKI Vol. 125
Pietrzynski, Gieren & Udalski 2003
Gonzalez et al.: bulge reddening maps and metallicity gradients 3
Fig. 2. J,H and Ks magnitudes di!erences between VVV catalogs and2MASS for stars. Dashed horizontal lines show the mean di!erencebetween both catalogs in the range denoted by the solid vertical lines
range 13>Kvvv >12 for all three bands were used as zero pointsto be applied to the magnitudes in the VVV catalogs in orderto obtain photometry fully consistent with 2MASS. The mag-nitude range used for the calibration is based on the fact that2MASS photometry is very accurate for our purpose down to amagnitude of K2mass !13 and therefore this range ensures thatthe di!erences are not a!ected by larger errors in the photom-etry or saturation. This procedure was carried on independentlyand zero points were calculated on each VVV tile catalog. Oncecalibrated, stars brighter than Ks!12 from 2MASS were used tocomplete the bright end of our final VVV catalog. These cata-logs, corrected from saturation and calibrated to 2MASS zeropoints were used for all the analysis. For this reason, we here-after refer to the usual J,H,Ks magnitudes corresponding to the2MASS photometric system. Figure 3 shows the final (J,J-Ks)CMD for a 20x20 arcmin region using the final catalog for fieldb278 as an example.
3. Relative reddening values
Extinction values are calculated following a procedure based inthe use of the colors of the bulge red clump (RC) to obtain red-dening values according to that of a reference field with knownextinction. In order to obtain the properties of a reference RCpopulation we have analyzed one of the fields in the spectro-scopic analysis of Zoccali et al. (2008). We have measured thecolor of the red clump in a 10x10 arcmin region centered inl=1.14 and b=-4.18 (hereafter Baade’s Window) for which wehave adopted a reddening value of E(B-V)=0.55 as in Zoccaliet al. (2008). Figure 4 shows the observed CMD for this regionand the (J-Ks) color histogram for which we find a mean value of(J-Ks)BW
RC =0.96 based on a Gaussian fit to the color distribution.By following this same procedure in any region of the Bulge, theextinction E(B-V) of that region can be related to that of Baade’sWindow by the following equation:
E(B " V) = E(B " V)BW " "(J " Ks)RC/(0.87 " 0.35) (1)
where E(B " V)BW is the extinction in Baades window and"(J " Ks)RC is the di!erence between the color of the RC inBaade’s Window (J-Ks)BW
RC and the observed color of the RC
Fig. 3. Final observed (J-Ks,Ks) CMD for the tile b278 once calibratedand completed with 2MASS photometry.
Fig. 4. Right panel shows the observed color magnitude diagram for a10x10 arcmin region centered in Baade’s Window and a selection boxcorresponding to the red clump. Left panel shows the color distributionof these red clump stars.
in each field (l,b) (J-Ks)lbRC as obtained from the Gaussian fit.
For our procedure we adopted the standard extinctions transfor-mations from Cardelli et al. (1989) of Ak = 0.35 · E(B " V),AH = 0.59 · E(B " V) and AK = 0.87 · E(B " V).
We are able to use this technique in all bulge fields coveredby the survey as VVV photometry reaches the RC even in thosehighly extincted regions. We have divided each VVV field intosmaller subfields of !6x6 arcmin in order to: i) Use small regionsin which we minimize the e!ects of di!erential reddening and ii)have enough stars in the subfield to populate the RC with goodenough number statistics to accurately trace the RC.
A fully de-reddened CMD of the field b306 is shown inFig 5. The e!ect of di!erential reddening is clearly observedand is consistent with the reddening vector, also shown in thefigure, for a 0.5 magnitude di!erence in E(B-V). We have car-
VVV Gle b278
Gonzalez et al. 2011
Red Clump in nearby galaxies in near-‐IR
Distance determinaGon: recipe 1. ExGncGon correcGon – select the reddening law
– Nishiyama et al. 2009: AK = 0.528 * [(J-‐Ks)0 – (J-‐Ks)] – Mean intrinsic color for RC in Baade’s Window (J-‐Ks)0 = 0.68
2. Make a luminosity funcGon and fit it with a 2nd order polynomial (underlying RGB) + a Gaussian (RC)
3. The peak of the Gaussian is the m(RC) 4. Distance modulus: (m – M)0,target = mΚ
RC – MΚRC – AΚ
MΚ
RC – RC zero point ΔMΚ
RC – populaGon correcGon with respect to populaGon used to establish the zero point Mλ
RC
Gonzalez et al.: Detection of the inner Galactic bar traced by the VVV survey
Fig. 1. K-band extinction map in Galactic coordinates showing the fields analyzed in this study. The 28 VVV tiles cover longitudes between!10" < l < 10" and latitudes !1.4" < b < !0.4", and 0.7" < b < 1.7".
2. The data
For this study we use the near-IR JK photometry from the VVVpublic survey1. A detailed description of the survey and the datacan be found in Minniti et al. (2010). In particular, the photo-metric calibration and dereddened magnitudes were calculatedas described in Gonzalez et al. (2011) and therefore are onlyshortly described here. Multiband catalogs were produced by acrossmatch of sources between single band catalogs producedat the Cambridge Astronomical Survey Unit (CASU) using theSTILTS code (Taylor 2006). For this work we analyzed a totalof 28 tiles, containing about 30 million measured sources, whichcover the survey region between !1.4" < b < !0.4" and +0.7" <b < +1.7" at longitudes between !10" < l < 10" (Fig.1). Thephotometric calibration was obtained by crossmatching sourcesof VVV with those of 2MASS flagged with best quality pho-tometry. This produced the final VVV J, H and Ks photometriccatalogs fully consistent with the 2MASS photometric system.The calibrated VVV tile catalogues were then dereddened andused to build the luminosity functions for each of the tiles.
3. Analysis and results
The extinction correction was done following the prescriptionsdescribed in Gonzalez et al. (2011). However, for the presentanalysis, we have adopted the Nishiyama et al. (2009) extinctionlaw, where Ak = 0.528 · E(J ! Ks) instead of the standard valuesfrom Cardelli et al. (1989), as the latter does not seem to be con-sistent with observations in the high reddening regions analyzedhere. Therefore, the dereddened Ks0 magnitudes were obtainedas:
Ks0 = Ks + 0.528[(J ! Ks)0 ! (J ! Ks)] (1)
where we use the mean intrinsic RC color for the bulge of(J!Ks)0 = 0.68 as measured in Baade’s Window (Gonzalez et al.2011). Note that by following the procedure in Gonzalez et al.(2011) we obtain reddening values in subfields of 6 # 6 arcmininstead of assigning individual extinction values to each star asin N05. Because of the e!ect of the broad bulge metallicity dis-tribution (e.g. Zoccali et al. 2008; Bensby et al. 2011), a singleintrinsic RC color cannot be adopted for all stars but to the meanof the observed RC color distribution in small subfields wherethe e!ects of di!erential extinction are minimized. Therefore,our procedure allows us to obtain dereddened Ks magnitudeswhile keeping the imprints of the original bulge stellar popu-lation. Figure 1 shows the final extinction map and the field cov-erage for this study.
We applied a color cut (J !Ks)0 > 0.4 to select the region ofthe color-magnitude diagram (CMD) dominated by bulge giants
1 http://vvvsurvey.org
and avoiding contamination from disk stars. Using the dered-dened Ks0 magnitudes of bulge giants we then built the luminos-ity function for each tile and followed the method introduced byStanek et al. (1996) to measure the RC mean magnitudes in or-der to trace the mean distance of the bulge population. The baseof the luminosity function was fitted with a second order poly-nomial and a Gaussian fit was then applied to measure the RCmean magnitude. Therefore the final fitting function is describedby:
N(Ks0 ) = a + bKs0 + cK2s0+
NRC
!RC$
2"exp! (KRC
s0! Ks0 )2
2!2RC
"(2)
The results of the Gaussian fits for each RC are listed inTable 1. Two main features in our procedure di!er from previoussimilar studies of the Galactic bar using RC giants. As describedbefore, we do not apply individual extinction corrections to eachstar. Instead, we attempt to maintain the color properties of bulgestellar population by obtaining reddening values in small sub-fields. Also, we include an additional Gaussian fit for a peak inthe luminosity function, distinct from the RC, which is centeredat Ks0 % 13.7. N05 mention the detection of this additional peakat a dereddened magnitude K0 %13.5 but they only performedGaussian fits up to a magnitude where the RC is una!ected bythis peak. Nataf et al. (2011) presented its detection, based onoptical OGLE data, and identified it as an anomalous red giantbranch bump (RGBb). In Gonzalez et al. (2011) the peak wasobserved using the VVV photometry up to b = !4". For largerdistances from the plane the detection of this peak becomes dif-ficult due to the split of the RC (McWilliam & Zoccali 2010).However we point out that, while the magnitude of the RC be-comes brigther toward positive longitudes according to the barorientation, the Ks0 magnitude of this additional peak changes inthe opposite direction. This is not expected if such feature cor-responds to the RGBb. A possibility could be that it instead cor-responds to RC stars from a more distant population. Assumingan intrinsic RC magnitude of MK = !1.55, the observed fainterRC peak would be at % 11.2 kpc from the Sun. Further discus-sion on this point, such as the number counts of the RC with re-spect to those of the secondary peak cannot be addressed in thiswork due to possible incompleteness at the magnitudes wherethis peak is observed, in particular for regions a!ected by highreddening and crowding. However, we assume here that the RCmean magnitudes remain reliable as long as the additional peakis included in the fitting procedure, independently of its origin.Figure 2 shows the results for the fits to the Ks,0 distribution ofRC stars in representative fields of the regions studied in thiswork.
Figure 3 shows the measured location of the Galactic baras obtained from the dereddened Ks0 magnitude of the RC atb = ±1", compared with those of N05 for b = +1". For this we
2
Tile b313l=-9.3
Tile b341l=-9.3
Tile b317l=-3.4
Tile b345l=-3.4
Tile b320l=+1.0
Tile b348l=+1.0
Tile b322l=+3.9
Tile b350l=+3.9
Tile b326l=+9.7
Tile b354l=+9.7
Gonzalez et al. 2011, A&A LeF
Red Clump magnitude variaGon as a funcGon of longitude at b=+1, -‐1 deg Stars are closer to us at posiGve longitude
K-‐band RC magnitude
2003MNRAS.343..539P
Percival & Salaris 2003
Alves 2000; Alves et al. 2002 – 2MASS & CIO + Hipparcos (Solar Neighb.) Grocholski & Sarajedini 2002 – WIYN Open clusters Pietrzynski et al. 2003 – LMC, SMC, Fornax, Carina (Araucaria project) Salaris & Girardi 2002 – populaGon effects (theoreGcal) Percival & Salaris 2003 – populaGon effects (empirical + models) Laney et al. 2012 – SAAO + Hipparcos (Solar Neighbourhood)
Oscar’s Bulge metallicity map
b
l
MK=-‐1.55 (solar metallicity, 10 Gyr old isochrone: Pietrinferni et al. 2004)
From RC distances to structure
Gonzalez et al. 2011, A&A LeF
Inner bar flaFening
Model: MarGnez-‐Valpuesta & Gerhard 2011 Gerhard & MarGnez-‐Valpuesta 2012
And below the plane…
Gonzalez et al. 2012, submiFed Gerhard & MarGnez-‐Valpuesta 2012
At b=-‐5, ~750 pc below the plane the bar flaFens much less
Second clump Tile b313l=-9.3
Tile b341l=-9.3
Tile b317l=-3.4
Tile b345l=-3.4
Tile b320l=+1.0
Tile b348l=+1.0
Tile b322l=+3.9
Tile b350l=+3.9
Tile b326l=+9.7
Tile b354l=+9.7
X (kpc)
Y (kpc)
K=11 K=11.5
K=12
K=12.5
K=13
K=13.5
K=14
1 kpc 2 kpc 3 kpc 4 kpc
SGll some head scratching …
11 12 13 14 15
-200
0
200
400
600
800
1 2
11 12 13 14 15
-200
-100
0
100
Re
sid
ual
XGAUSSFIT
Baseline Coefficients
-42.015479
Peak Center FWHM
1 762.67531 12.616795 0.71169916
2 384.62215 13.323637 1.2156681
Ks
Tile: b255 l = 8, b=-‐6.5 Second clump ~2.5 kpc behind the bar and ~ 1kpc below the plane!
0.2
0.4
0.6
0.8
1.0
1.2
0.2
0.4
0.6
0.8
1.0
1.2
0.2
0.4
0.6
0.8
1.0
1.2
0.2
0.4
0.6
0.8
1.0
1.2
0.2
0.4
0.6
0.8
1.0
1.2
11 12 13 14 15
0.2
0.4
0.6
0.8
1.0
1.2
Ks
l=+10
l=+4
l=+1
l=-‐1
l=-‐4
l=-‐10
Summary
• VVV depth and resoluGon provides ideal dataset for inner Bulge structure studies
• RC, eclipsing binaries, RR Lyr à tracing different components?
• Complexity: – patchy exGncGon (BEAM calculator) – Mix of populaGons – bar/bulge, thin disk and spiral arms, thick disk?
– Comparison with models