On the origin of the galaxy star-formation-rate sequence: evolution and scatter

We use a semi-analytic model for disk galaxies to explore the origin of the time evolution and small scatter of the galaxy SFR sequence - the tight correlation between star formation rate (SFR) and stellar mass (M-star). The steep decline of SFR from z similar to 2 to the present, at fixed M-star, is a consequence of the following. First, disk galaxies are in a steady state with the SFR following the net (i.e. inflow minus outflow) gas accretion rate. The evolution of the SFR sequence is determined by evolution in the cosmological specific accretion rates, alpha (1 + z)(2.25), but is found to be independent of feedback. Although feedback determines the outflow rates, it shifts galaxies along the SFR sequence, leaving its zero-point invariant. Second, the conversion of accretion rate to SFR is materialized through gas density, not gas mass. Although the model SFR is an increasing function of both gas mass fraction and gas density, only the gas densities are predicted to evolve significantly with redshift. Third, star formation is fueled by molecular gas. Since the molecular gas fraction increases monotonically with increasing gas density, the model predicts strong evolution in the molecular gas fractions, increasing by an order of magnitude from z = 0 to z similar to 2. On the other hand, the model predicts that the effective surface density of atomic gas is similar to 10M(circle dot) pc(-2), independent of redshift, stellar mass or feedback. Our model suggests that the scatter in the SFR sequence reflects variations in the gas accretion history, and thus is insensitive to stellar mass, redshift or feedback. The large scatter in halo spin contributes negligibly, because it scatters galaxies along the SFR sequence. An observational consequence of this is that the scatter in the SFR sequence is independent of the size (both stellar and gaseous) of galaxy disks.