The star formation history as a function of type: constraints from galaxy counts

Abstract

Deep galaxy counts are among the best constraints on the cosmic star formation history (SFH) of galaxies. The evolution of the star formation activity is followed on a wide range of redshifts (0< z < 4) covering most of the history of the Universe. Two incompatible interpretations of the observations are currently competing. After applying star formation rate (SFR) conversion factors to the CFRS, Halpha or ISO samples, many authors conclude to a strong increase (~ a factor 10) of the SFR from z=0 to z=1. They also find some evidence for a peak at z~1 and for a rapid decrease at higher redshifts. On the other side, the Hawaii deep surveys favor only a mild increase between z=0 and 1 (Cowie et al., 1996, 1999). To understand the reason for these discrepant interpretations, we consider three classes of galaxies: E/S0, Sa-Sb-Sbc, Sc-Sd-Im and bursting dwarfs. We use the new version of our evolutionary synthesis code, PEGASE which takes into account metallicity and dust effects. The main results are: i) Late-type galaxies contribute significantly to the local SFR, especially bursting dwarfs (Fioc and Rocca-Volmerange, 1999). Because of that, the cosmic SFR can not decrease by a factor 10 from z=0 to 1. This is in agreement with Cowie's result. ii) The SFR of intermediate-type galaxies has strongly decreased since z=1. Though the decrease is less than what find Lilly et al., 1996, this suggests that the CFRS and Halpha surveys are dominated by such bright early spirals. iii) The contribution of early-type galaxies increases rapidly from z=1 to their redshift of formation (> 2-3 for cosmological reasons). Their intense star formation rates at high-z give strong constraints on early ionization phases, primeval populations or metal enrichments.(abridged)