The clustering of simulated quasars

Abstract

We analyze the clustering properties of quasars simulated using a semianalytic model built on the Millennium Simulation, with the goal of testing scenarios in which black hole accretion and quasar activity are triggered by galaxy mergers. When we select quasars with luminosities in the range accessible by current observations, we find that predicted values for the redshift evolution of the quasar bias agree rather well with the available data and the clustering strength depends only weakly on luminosity. This is independent of the lightcurve model assumed, since bright quasars are black holes accreting close to the Eddington limit. We also used the large catalogues of haloes available for the Millennium Simulation to test whether recently merged haloes exhibit a stronger large-scale clustering than the typical haloes of the same mass. This effect might help to explain the very high clustering strength observed for z~4 quasars. However, we do not detect any significant excess bias for the clustering of merger remnants, suggesting that objects of merger-driven nature do not cluster significantly differently than other objects of the same characteristic mass.