Post on 19-Dec-2015
Supernova Legacy Survey
Cosmology, Spectroscopy, and Progenitors
T. Justin Bronder, DPhil Candidate Oxford University
Isobel Hook, Supervisor
Overview• Background on ‘Type Ia Cosmology’• Supernovae Legacy Survey Overview
– Goals– Methods
• Photometry and Spectroscopy
– Year 1 Cosmology Results
• Quantitative Spectroscopy of Hi-z SNe Ia– Goals / Methods / Results
• Brief intro to other SNLS Science – Ex) SNe Ia Hosts and Delay Times– Future work: Quantify evolution? Constrain progenitors?
Background
Hydrogen? No H2 present: H2 present:
General group? Type I supernovae Type II supernovae
Other properties? Si He neither Photometry/spectral properties
Specific type? Ia Ib Ic IIp IIn II...
- defining Type Ia Supernovae
Background
Hydrogen? No H2 present: H2 present:
General group? Type I supernovae Type II supernovae
Other properties? Si He neither Photometry/spectral properties
Specific type? Ia Ib Ic IIp IIn II...
- defining Type Ia Supernovae
Background (cont)
• Cosmological utility as ‘standardizeable candle’– Variance in B’ peak magnitude – Peak mag. correlates w/
decline rate of light curve – empirical correction reduces peak magnitude deviations enough for cosmological use
– Chart SNe Ia with z to determine matter / dark matter content in the Universe
‘s’ parameter – Goldhaber et al 2001
Cosmological constraints come from many sources
Combine with Type Ia supernova surveys = ‘Cosmological Concordance Model’
Results inconsistent with M=1 spatially flat cosmology
(SNe too faint)
SN data favor >0
What is dark energy?
Differentiate via the equation of state
<w = p
Background (cont)
• Issues / Concerns / Criticisms– Understanding of SNe Ia physics
• Extensive agreement (theory / observation / models) on basic model:– C/O White Dwarf exceeds MCh
– Thermonuclear runaway disrupts WD, late time – radioactive decay of unstable nuclear burning products
• Many questions follow:– Progenitor scenario – Single or Double Degenerate
– Burning front – Detonation / Deflagration / Both
– ‘homogeneity’? Case of 1991T, 1991bg
– Host dependence? Early galaxies host dimmer SNe Ia
– Type Ia Cosmology Statistics and Systematics• Luminosity calibration• Evolution• Extinction
SNLS collaboration http://cfht.hawaii.edu/SNLS/
Chris Pritchet: U. VictoriaRay Carlberg: U. Toronto Andy Howell: U. TorontoMark Sullivan: U. TorontoArif Babul: U. VictoriaDavid Balam: U. Victoria Sara Ellison: U. Victoria F.D.A. Hartwick: U. Victoria Henk Hoekstra: CITADon Neill: U. TorontoJulio Navarro: U. Victoria Kathy Perrett: U. Toronto David Schade: HIAPierre Astier : CNRS-IN2P3, Paris Eric Aubourg Christophe Balland
Luc Simard: HIA Peter Stetson: HIA Sidney van den Bergh: HIA Jon Willis: U. Victoria Isobel Hook: U. Oxford Justin Bronder: U. OxfordRichard McMahon: U. CambridgeReynald Pain: CNRS-IN2P3, Paris Saul Perlmutter:LBNLRobert Knop: U. VanderbiltJames Rich: CEA-Saclay Nic Walton: U. CambridgeEric Smith: Vanderbilt UniversityGreg Aldering: LBNL Lifan Wang: LBNL Rachel Gibbons: LBNL Vitaly Fadayev: LBNL
Stephane Basa Sylvain Baumont Sebastien Fabbro Melanie Filliol Ariel Goobar: StockholmDelphine Guide Julien Guy Delphine Hardin Nicolas Regnault Tony Spadafora: LBNLMax Scherzer: LBNL Harish Agarwal: LBNLHerve Lafoux Vincent Lebrun Martine Mouchet Ana Mourao Nathalie Palanques Gregory Sainton
Canada, France, UK, US, Sweden, Portugal
I – SNLS Overview• Populate Hubble Diagram with over ~700 Type Ia SNe (.2 <z<
1.0) to estimate w to + 0.1• CFHT ‘Megacam’ used to acquire multiple (~5 epochs monthly)
points in g’r’i’z’ for 1- potential candidate ID and 2-luminosity calibration (‘stretch factor’)
• Candidates verified w/ 8m-class telescopes – Gemini, VLT, and Keck
• Publications:– astro-ph/0509195 - Gemini spectroscopy, Howell et al– astro-ph/0510447 - Measurements for Cosmology, Astier et al– ‘Photometric Selection of high-z Type Ia Candidates’ – Sullivan et al– VLT Spectroscopy summary – in prep– Much more on Hi-z SNe science in the works
-Possible Type Ia candidates followed up with 8m spectroscopy main purpose is candidate ID and redshift confirmation
-GMOS: 0.75’’ slit range = 465-930 nm 1.34 A/binned pixel disp. 45-90 min exposure time
-FORS1: 0.7’’-1.3’’ slit range = 445-1100 nm 2.69 A/binned pixel disp. 25-62.5 min exposure time
Smoothed spectrum allowing for:• template host galaxy subtraction• Reddening
Extracted spectrum
SiII
- latest LSS / CMB / Baryon Acoustic Peak Oscillations results added to concordance model
(Allen, Schmidt, & Fabian 2002 / Spergel et al 2003 / Eisenstein et al 2005)
- SNLS Hubble diagram – Astier et al (accepted by A&A)
- M = .263 + .042 (stat) + .032 (sys) (flat CDM model)
- w = -1.023 + .090 (stat) + .054 (sys)
I – SNLS Overview• Spectroscopy implications
– Only 1 epoch• Velocity gradients, etc. not viable
– SNR to maximize number of candidates observed + identified• No synthetic spectra/detailed line analysis• SNR-driven error bars on any results
• Well suited for SNLS purposes– Candidate ID via template-matching
• ‘Superfit’ (Howell et al. 2002, Lidman et al. 2005)• Results in Gemini/VLT data papers• 72 observations - 47 confirmed Ia (Gemini, 12 months)• 108 observations – 67 confirmed Ia (VLT, 18 months
II – SNLS Spectroscopic Science• Main purpose of spectra is ID/z
– More thorough analysis a necessity for• Object clarification / independent identification
• Physical insight – Evolution / systematic checks / progenitors
– Needs + survey constraints imply data set is best suited for:• Quantify the distribution of spectral properties (check for evolution w/ z, environment)
– Specific comparison to low-z Ia SNe population
– Type Ia sub-types ?
• Object clarification independent of 2 fits
• Another parameter space to explore the systematics of this large sample for cosmology
II – Science w/ SNLS spectroscopy - background
- expansion velocity of CaII H&K feature
-May also probe for Z effects (Hoflich et al 1998, Lentz et al 2000) if measured on a high-z population
II – Science w/ SNLS spectroscopy - background
-R{CaII} and R{SiII} – spectral feature ratios from Nugent et al.1995
-Utilized in other empirical treatments of Type Ia spectra
II – Science w/ SNLS spectroscopy - background
- Velocity gradients / spectral feature ratios explored empirically in Benetti et al 2004
- continuum of Type Ia properties? Two different populations?
II – Science with SNLS spectroscopy
-Equivalent Widths (EW)
-Shape independent method of spectral feature strength
-Folatelli 2004 – EW measurements for
Type Ia – specific features
-Folatelli found this measurement useful for quantifying SNe Ia spectral -homogeneity – subtype ID, correlations w/ lightcurve-shape params
II- EW Results – Low z• Distribution of spectral properties
» EW {CaII} - Check of overluminous SNe Ia
II- EW Results – Low z• Distribution of spectral properties
» EW {SiII} - smaller epoch evolution than other features
- correlates to Lightcurve parameters
II- EW Results – Low z• Distribution of spectral properties
» EW {SiII} - smaller epoch evolution than other features
- correlates to Lightcurve parameters
II- EW Results – Low z• Distribution of spectral properties
» EW {MgII} - clear check of underluminous objects
- additional correlation to
luminosity params
III – Equivalent Width results
• Quantitative analysis– Extensive low-z data employed to generate a mean trend
for branch normal SNe Ia– Comparison ‘model’ for high-z objects
• Ejection velocities of CaII H&K feature also measured– Gaussian fit to feature – minima of fit = blueshift
velocity– Low redshift branch normal ‘model’ computed for
quantitative comparison
III - Results• Distribution of spectral properties :
» EW {CaII} - SNLS 1st Year - reduced chi squared: .511
III - Results• Distribution of spectral properties :
» EW {CaII} - epoch and sampling ‘free’ comparison via residuals – compare SNe populations w/ K-S Test – no significant difference
III - Results• Distribution of spectral properties :
» EW {SiII} - SNLS 1st Year - reduced chi squared: .471
III - Results• Distribution of spectral properties :
» EW {SiII} - KS test results show a difference (not significant) – identifies outliers independently of photometry / typing
III - Results• Distribution of spectral properties :
» EW {SiII} - recall light curve – spectra correlation
III - Results• Distribution of spectral properties :
» EW {SiII} - most objects follow Low-z trend… shift in distribution? Significant outliers?
III - Results• Distribution of spectral properties :
» EW {MgII} - SNLS 1st Year - reduced chi squared: .421
III - Results• Distribution of spectral properties :
» EW {MgII} - K-S test results similar to EW{CaII} – note possible overluminous outliers
III - Results• Distribution of spectral properties :
» Vej {CaII} - SNLS 1st Year - reduced chi squared: .372
III - Results• Distribution of spectral properties :
» Vej {CaII} - Residuals – outliers noted – trend or small numbers? Possible
metallicity indication – epoch not quite early enough to match predictions
III - Results• Quantitative look at distribution of these spectral properties shows ‘no
significant difference’ at high-z– Caveat – large error bars test for broad consistency
– K-S tests on residuals support 2 results – no major / systematic shift in Ia properties at high-z– Definite differences
• Are these expected?• Implications for SNe Ia cosmology?
• Object sub-typing / systematic checks / PG’s – Spectroscopic outliers noted for i) follow up investigation ii)
‘spectroscopic’ SNLS Hubble diagram to check cosmology systematics– Host Galaxy – Luminosity/sub-type correlation at low z
• Quantitative spectroscopy + high-z host/environment data will enable exploration of this observation test age/metallicity and progenitor dependencies
IV – SNLS science [brief intro]- Sullivan et al (2003) – ‘morphological’ SNe Ia Hubble diagram
- Residuals for objects in E/S0 hosts were smaller than other host types
- results w/in each type still supported dark matter model
- extrapolate to higher redshift – unveil population / age / evolutionary differences clues to progenitors?
… delay times?
IV – SNLS science [brief intro]- delay times useful in constraining PG scenarios (SD v DD)
- Previous work (Madau et al 1998, Dahlen & Fransson 1999, Gal-Yam & Maoz 2004, Strolger et al
2004) gave delay times from ~ 1.0 to 4.0 Gyr
- heavily dependent on SFR assumptions
- no results quite explained Fe / O (Ia / CC SNe) ratio in galaxy clusters
- Scannapieco & Bildsten 2005 – 2-component Type Ia formationScannapieco & Bildsten 2005 – 2-component Type Ia formation
i) ‘prompt’ – proportional to SFR ii) extended – proportional to mass
- quantifies simple observation that SNe Ia are seen in all galaxy types
- SNLS data: Host spectra can be used to quantify local SFR and mass = specific star-formation rate test this model
- initial results ‘agree’ – two Type Ia channels or one channel with a large range in PG system age
1
o10
o10 Gyr 10
.10
.SNRateM
SFRB
M
MAt
IV – SNLS science [brief intro]- other evidence for two-channel/extended PG time Ia channels?
- distribution of light-curve properties (here presented by star-formation rate rather than morphology courtesy of M. Sullivan) corroborates previous observations
- evidence for two PG channels or populations?
IV – SNLS science [brief intro]- evidence for two PG channels or populations?
- additional physical insight with spectroscopy?
IV – SNLS science [brief intro]- evidence for two PG channels or populations?
- additional physical insight with spectroscopy?
Conclusion• Brief summary of ‘Type Ia Cosmology’
– Earlier results (Riess et al 1998, Perlmutter et al 1999) supported by SNLS– Will also constrain w for additional cosmological insight
• SNLS Science and Spectroscopy– Successful at main goal – object ID / redshift– Quantified analysis can also i – check for broad consistency to low-z
population ii – id sub-types and outliers exert an additional systematic control on large SNLS data set iii – combine with other SNLS science to provide insight into Type Ia physics / pg’s