Evolution of Restframe Luminosity Density to z2 in

Evolution of Rest-frame Luminosity Density to z=2 in the GOODS-S Field Tomas Dahlen, Bahram Mobasher, Rachel Somerville, Lexi Moustakas Mark Dickinson, Harry Ferguson, Mauro Giavalisco and the GOODS Team GOODS-S multi-waveband observations Photometric redshifts Results on LFs and luminosity densities Summary ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

Two sets of data covering GOODS-S WFI R-selected CTIO-U, WFI-BVRI, FORS-RI, ISAAC JHKs, SOFI JHKs psf: 0. 9 arcsec Area: 1100 sq. arcmin Redshift range: 0. 1<z<1. 0 RAB < 24. 5 Number of galaxies: 18300 Rest-frame U and B to z=1 ISAAC Ks-selected ACS BViz, ISAAC JHKs psf: 0. 45 arcsec Area: 130 sq arcmin Redshift range 0. 1<z<2. 0 Ks. AB<23. 2 Number of galaxies: 2768 Rest-frame J to z=1 and B to z=2 ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher ISAAC now twice as large

GOODS Phot-z Code / six templates used / Luminosity function used as prior / Cosmic opacity from Madau et al. / Extinction allowed as a free parameter, estimating E(B-V) for each galaxy / Interpolates the spectral types / Easily extended to other bands (ie IRAC, GALEX etc)

Luminosity functions are calculated using * 1/Vmax method * Maximum likelihood method Traditionally (ideally): each galaxy has one redshift -> one absolute magnitude -> one galaxy added to magnitude bin in LF Using phot-z's: Phot-z's have relatively large errors Each galaxy is represented by a redshift distribution

Photometric Redshifts Template fitting method Input LF prior Extinction correction Comparison with ~400 spec-z's: Accuracy: s(z)/(1+zspec) = 0. 12 Accuracy. s(z)/(1+zspec) = 0. 06, after excluding 3% out-liers ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

Photometric Redshifts & the Luminosity Function We use the 1/Vmax-method to calculate the LF: F(M)d. M = S 1/Vi(Mi) Uncertainties in phot-z may introduce bias when determining the LF (Chen et al. 2003) Instead of a single redshift, we use a probability distribution for each object derived from phot-z method. The 1/Vmax-method is now expressed as: F(M)d. M = SPi/Vi(Mi) ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

Results: The Luminosity Function, 0. 1<z<0. 5 Ho=70 km/s/Mpc, WM=0. 3, WL=0. 7 'Quasi-local' B-band LF for different types Late-types follow composite LF Early-types have Gaussian shape Starbursts are fainter ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

Results: The Luminosity Function, 0. 1<z<0. 5 'Quasi-local' U-band LF for different types Similar to B-band but Starbursts more dominating ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

Results: The Luminosity Function, 0. 1<z<0. 5 'Quasi-local' J-band LF for different types Steep faint-end slope (as optical) Starbursts less abundant Band M* a f* U -20. 1 -1. 33 0. 0026 B -21. 4 -1. 40 0. 0027 J -23. 6 -1. 48 0. 0008 ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

Results: The Luminosity Function, 0. 1<z<1. 0 U-band B-band J-band ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

Results: The Luminosity Function, 0. 1<z<1. 0 No strong evolution in faint-end slope in U, B or J Brightning of M* in U and B to z~1 (DM~0. 5 mag) Slight fading of M* in J to z~1 Faint-end slope a~-1. 3 - -1. 4, in all bands U&B slopes consistent with e. g. , COMBO-17, FORS Deep Field J-band consistent with local 2 MASS K-band slope a=-1. 09? Yes! ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

Results: Luminosity density evolution The LF is integrated to derive luminosity density rn = G(2+a)f*L* (for Schecheter function approximation) Mean increase in all bands Largest evolution in U-band Relative contribution to rn from different spectral types: Significant increase of Starburst contribution with z Highest fraction Starburst in U-band Decrease of Early-type fraction with z Lowest Early-type fraction in U-band ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

B-band luminosity density evolution to z=2 R- and K-selected measurements consistent where they overlap Increase by factor ~1. 7 in r. B to z=1 (compared to local value), Increase consistent with e. g. , COMBO-17 No significant evolution at z>0. 6 (in any survey) ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher

Conclusions Important to incorporate phot-z errors when deriving the LF No significant evolution in the shape (a) of the LF to z=1 in U, B, and J Brigthening of M* in optical bands (U, B) with redshift Indication of fading M* with redshift in J Faint-end slope a~-1. 3 - -1. 4 to z~1 Different shape of type-specific LFs (Early-types, Late-types and Starbursts) Increase in U-band luminosity density to z=1 (less significant in B and J) Sharp increase of the relative contribution from Starbursts in the luminosity density to z=1, especially in U, along with a decrease in the Early-type contribution The B-band luminosity density increases by a factor ~1. 7 to z~1. No significant evolution in the B-band luminosity density at z > 0. 6 ”Massive Galaxies”, STSc. I, 27 Sept 2004 Bahram Mobasher