Determining the Properties and Evolution of Red Galaxies from the Quasar Luminosity Function

We study the link between quasars and the red galaxy population using a model for the self-regulated growth of supermassive black holes in mergers involving gas-rich galaxies. In this picture, mergers drive nuclear inflows of gas, fueling starbursts and obscured quasars until feedback energy from black hole growth expels the surrounding gas, rendering the quasar briefly visible as a bright optical source. The quasar dies when there is no longer a significant supply of gas to power accretion, and the stellar remnant relaxes as a passively evolving spheroid satisfying the MBH −σ relation and lying on the fundamental plane. The same process that halts black hole growth also terminates star formation in the remnant, accounting for the observed red galaxy population in the bimodal color/morphology distribution of galaxies. Using a model for quasar lifetimes and evolution motivated by hydrodynamical simulations of galaxy mergers, we de-convolve the observed quasar luminosity function at various redshifts to determine the rate of formation of black holes of a given final mass. Identifying quasar activity with the formation of spheroids in the framework of the merger hypothesis, this enables us to deduce the corresponding rate of formation of spheroids with given properties as a function of redshift. This allows us to predict, for the red galaxy population, the distribution of galaxy velocity dispersions, the galaxy mass function, mass density, and star formation rates, the luminosity function in many observed wavebands (e.g., NUV, U, B, V, R, r, I, J, H, K), the total number density and luminosity density of red galaxies, the distribution of colors as a function of magnitude for several different wavebands, the distribution of colors as a function of velocity dispersion, the distribution of mass to light ratios as a function of mass, the luminositysize relations, and the typical ages and distribution of ages (formation redshifts) as a function of both mass and luminosity. For each of these quantities, we predict the evolution from redshift z = 0−6. Each of our predictions agrees well with existing observations, without the addition of tunable parameters; the essential observational constraints and inputs come from the the observed quasar luminosity function. We further demonstrate that these predictions are skewed by several orders of magnitude if we adopt simpler, traditional models of quasar lifetimes in which these objects turn “on/off” or follow simple exponential light curves, instead of the more complicated quasar evolution implied by our simulations. Subject headings: quasars: general — galaxies: nuclei — galaxies: active — galaxies: evolution — cosmology: theory