Spitzer view on galaxy evolution in the last

Spitzer view on galaxy evolution in the last 12 Gyr (since z~4) Pablo G. Pérez-González, MIPS Instrument Team UCM IAC Seminar January 23, 2007 Uof. A Astrid

IAC Seminar January 23, 2007 Hopkins & Beacom 2006, Ap. J, 651, 142 Pérez-González, et al. , 2005, Ap. J, 630, 82 How stars formed: Lilly-Madau plot

Two Questions on galaxy evolution Are SFR estimations (and extinctions) good? Also get better masses, (metallicities, …) How GALAXIES formed and evolve? Link SFR to stellar mass (and metallicity). IAC Seminar January 23, 2007

Getting better SFRs M 51 M 81 24 -8. 0 -3. 6 mm 8. 0 -5. 8 -4. 5 -3. 6 mm IAC Seminar January 23, 2007

Improving models Pérez-González, et al. , 2006, Ap. J, 648, 987 (and errata) Pérez-González, et al. , 2007, in preparation Consistent models of the UV-to-FIR emission (energy balance). IAC Seminar Stars (several populations), (hot) gas & dust (several components). January 23, 2007

Getting better SFRs Pérez-González, et al. , 2006, Ap. J, 648, 987 (and errata) (Calzetti, et al. , 2005) m=1. 3 (Calzetti, et al. , 2005) m=1. 0! Best SFR estimations? : add ‘observed’ and ‘obscured’ SFRs. L(24) vs. SFR (better than TIR) calibration extended to LIRGs (Alonso-Herrero et al. 06), low Z regions (Relaño et al. 07), and SINGS IAC Seminar galaxies (Kennicutt et al. 2007, Prescott 2007). January et 23, al. 2007

Getting better SFRs Pérez-González, et al. , 2006, Ap. J, 648, 987 (and errata) Scatter~35% TIR better correlated with UV/optical (colder dust –dominating L(TIR)- heated by older stars –Bendo et al. ‘ 06, Ap. J, 652, 283 -). And 24 mm is better correlated with Ha (hotter dust is heated by IAC Seminar more energetic photons or trapped. January emission-line photons). 23, 2007

Application to high redshift Combination of IR Spitzer data and Ha observations to get better SFRs and extinction properties (and metallicities) for z~2 galaxies: 4 nights on Gemini-S+GNIRS in Nov ‘ 06 3 nights on VLT+ISAAC in Dec ’ 06 applied for Subaru+MOIRCS (all sources!!) systematic survey with F-2&EMIR/GOYA Similar work for z~1 sources (PI Egami): Subaru+MOIRCS observations in 07 A. IAC Seminar January 23, 2007

MIPS GTO/CS database Survey Chandra GALEX EIS GOODS GEMS GOODS COMBO 17 MUSIC LCS Spitzer VLA VVDS/FORS 2 CDFS Bands hard, soft FUV, NUV UBVRIJK bviz vi JHK 17 filters 9 filters RIz IRAC MIPS 6 cm, 20 cm spectra IAC Seminar January 23, 2007 Comments merged no images

MIPS GTO/CS database Survey Chandra Subaru GOODS UCM Spitzer VLA TKRS/Cowie HDFN Bands hard, soft UBVRIz “HK” bviz JK IRAC MIPS 6 cm, 20 cm spectra IAC Seminar January 23, 2007 Comments merged no images

MIPS GTO/CS database Samples in CDFS, HDFN, Lockman Hole, Extended Groth Strip, ELAIS, and XMM Deep Field (CATALOGS & IMAGES, if available). 80, 000+ IRAC sources (50, 000 already with merged photometry) in ~3, 000 arcmin 2: mass selected sample (at 3. 6 -4. 5 mm) additional SFR selection (MIPS fluxes and UV emission from SEDs). Merged photometry in 25 -60 bands (including spectroscopy –redshifts-). Photo-z’s and stellar masses estimated with stellar population models applied to each source. IAC Seminar January 23, 2007

NIR 1 stellar population + gas + dust IRAC GALEX (SFR~e-t/t, SALP IMF, CF 00) IAC Seminar January 23, 2007 (CE 00) Pérez-González, et al. , 2007, Ap. J (in prep. ) Optical MIPS Pérez-González, et al. , 2005, Ap. J, 630, 82 “rainbow” catalogs

IAC Seminar January 23, 2007 Pérez-González, et al. , 2007, Ap. J (in prep. ) Pérez-González, et al. , 2005, Ap. J, 630, 82 “rainbow” catalogs

IAC Seminar January 23, 2007 Pérez-González, et al. , 2007, Ap. J (in prep. ) Pérez-González, et al. , 2005, Ap. J, 630, 82 “rainbow” catalogs

Photo-z: state of the art Optical/NIR based photo-z techniques Hyper-z: BC 03, SB 99 stellar templates Bayesian: improved model-data comparison, priors, less outliers Photo-z’s adding Spitzer data (UV-to-FIR) No complete set of models available (optical through FIR): Devriendt et al. 1999, Dale et al. 2004, Chary & Elbaz 2001 Empirically built templates (~2000 galaxies) Neural networks IAC Seminar January 23, 2007

Photo-z vs. spectroscopy HDFN Success: 88% excellent 95% good < z>=0. 003 <sz/(1+z)>=0. 031 P. -G. , et al. , 2005, Ap. J, 630, 82 IAC Seminar P. -G. , et al. , 2007, Ap. J (in prep. ) January 23, 2007

Photo-z vs. spectroscopy CDFS Success: 85% excellent 93% good < z>=0. 007 <sz/(1+z)>=0. 038 P. -G. , et al. , 2005, Ap. J, 630, 82 IAC Seminar P. -G. , et al. , 2007, Ap. J (in prep. ) January 23, 2007

Redshift distribution Large effect of cosmic variance SFR selected sample with MIPS/UV 6, 000+ sources at 1<z<2 3, 000+ sources at 2<z<4 15, 000+ sources at 0<z<1 IAC Seminar January 23, 2007 redshift Pérez-González, et al. , 2007, Ap. J (in prep. ) Mass selected sample up to z~2 (and beyond) Pérez-González, et al. , 2005, Ap. J, 630, 82 PAHs exist at z>1!!

Stellar masses P. -G. , et al. , 2007, Ap. J (in prep. ) IAC Seminar redshift January 23, 2007

Most massive galaxies are bright at IR at high-z Very massive galaxies (M>1012. 0 Mʘ) formed at z~3. Galaxies with 1011. 5<M<1012. 0 Mʘ formed at z>1. SFR density dominated by less IAC massive Seminar systems at z<1. January 23, 2007 Pérez-González, et al. , 2007, Ap. J (in prep. ) Downsizing scenario Pérez-González, et al. , 2005, Ap. J, 630, 82 Evolution of the SMF

More proofs of Downsizing P. -G. , et al. , 2007, Ap. J (in prep. ) IAC Seminar January 23, 2007

IAC Seminar January 23, 2007 Pérez-González, et al. , 2007, Ap. J (in prep. ) Pérez-González, et al. , 2005, Ap. J, 630, 82 Evolution of the SM density

Most stellar mass (>90%) was formed after z~4 (12 Gyr). The Universe formed half of its stellar mass before z~1. Faster evolution (massive galaxies) first (0. 07 Mʘyr-1 Mpc-3). -1 -3 IACMSeminar Milder evolution after z~1 (0. 03 ʘyr Mpc ). January 23, 2007 Pérez-González, et al. , 2007, Ap. J (in prep. ) Pérez-González, et al. , 2005, Ap. J, 630, 82 Evolution of the SM density

Summary & Conclusions Two research paths to improve our knowledge about the evolution of galaxies: combining samples & data at different l’s (better SFRs, masses, …). Relating the cosmic SFR density to the stellar mass density and the metallicity enrichment history. Best SFR estimates by combining data probing the unobscured SF (Ha, UV) and the extincted SF (MIR or FIR). See, e. g. , Bell 2003, Iglesias-Páramo et al. 2005, Hopkins & Beacom 2006. L(24) is the best estimator [better than L(TIR)] of Labs(Ha). Our spectral modeling technique includes the 3 visible components of a galaxy: stars, gas, and dust. We consistently combine the 3 components to reproduce the observed SEDs. We have built a representative template set of 2000 SEDs covering the UV-FIR, used to obtain photo-z’s and stellar masses for 80, 000 sources (mass selected) in ~3000 arcmin 2. With this sample, we have calculated the stellar mass function for several redshift bins up to z~4. Galaxy formation presents a downsizing scenario: higher mass galaxies are formed first (they are older) in violent SF episodes, lower mass galaxy formation continues to low-z. IAC Seminar January 23, 2007

Index Madau plot Goals of the project Comparing SFR estimators M 81 and M 51 TIR Metallicity effects LIRGs and ULIRGs Cosmologial Surveys Modeling procedure Photometric redshifts Redshift distribution (mass and SFR selected) Mass limits Evolution of the stellar mass function Evolution of the stellar mass density Recovering the Lilly-Madau plot Summary and Conclusions IAC Seminar January 23, 2007

Metallicity effects L(24)/Lcorr(Ha) correlates with metallicity! But just for entire galaxies, not individual SF regions!! Leakage of Ly photons and properties of ISM (U)LIRGs dwarfs Relaño, et al. , 2007, in preparation IAC Seminar January 23, 2007

LIRGs and ULIRGs Alonso-Herrero, et al. , 2006, Ap. J, 650, 835 For LIRGs and ULIRGs, we can find an empirical correlation IACslope Seminar between L(24) and Lc(Ha), but the is very different from 1. January 23, 2007

Spitzer images (CDFS) 1’ K (ISAAC 400 ks) IAC Seminar January 23, 2007

Spitzer images (CDFS) 1’ IRAC (3. 6 mm 12 ks) IAC Seminar January 23, 2007

Spitzer images (CDFS) 1’ NB 816 (Subaru) IAC Seminar January 23, 2007

Spitzer images (CDFS) 1’ MIPS IAC Seminar January 23, 2007

Recovering the Lilly-Madau plot Pérez-González, et al. , 2005, Ap. J, 630, 82 Hopkins & Beacom 2006, Ap. J, 651, 142 IAC Seminar January 23, 2007