Modelling the cosmic evolution of the brightest quasar population

We develop and present models within a semi-analytic merger tree based framework for the evolution of the brightest quasar population from redshifts z ~ 6 to the present, based on Eddington limited accretion on to supermassive black holes (SMBHs) and driven by galaxy mergers. We first demonstrate that even the simplest possible model can uncover constraints on the nature of quasar and SMBH evolution due to the rapid growth required to produce the bright high redshift quasars. This model is then gradually built up by sequentially adding in simplistic but plausible recipes for gas cooling, star formation and various modes of feedback with the aim not of producing a single highly-tuned fit to the observed evolution but rather of isolating and understanding the effects of these processes upon such evolution. Our final models are able to successfully reproduce the basic observed evolution of the brightest quasars, from the initial rise to the subsequent turnover in quasar numbers and then the sharp drop in numbers at the lowest redshifts, while simultaneously matching the universal cold gas density and star formation rate. The areas where our simplistic recipes fail are highlighted and we suggest several ways in which these problems might be rectified. The overall picture which emerges from our work is one where early rapid growth of SMBHs in galaxies is eventually slowed by self-regulation where feedback from the central source drives out further fuel and sets up the M - s relation. As galaxies then join groups and clusters at still later redshifts, however, further significant accretion is instead prevented by a suppression of gas cooling which is driven by some other mode of feedback. We lastly present preliminary results that suggest direct radiative heating by the central source is not an effective mechanism for this later mode of feedback and offer some alternatives.