The Clustering of Galaxies around Quasars

We study the cross-correlation between quasars and galaxies by embedding models for the formation and evolution of the two populations in cosmological N-body simulations. We adopt the quasar evolution model of Kauffmann & Haehnelt (2000), in which supermassive black holes are formed and fuelled during major mergers. We define the “bias” parameter bQG as the ratio of the cross-correlation function ξQG to the galaxy auto-correlation function ξGG. On scales larger than 1 h−1 Mpc, the values of bQG predicted by our models at low redshift depend very little on galaxy selection. They measure the characteristic mass of the dark matter halos that host quasars and can be used to estimate the typical quasar lifetime. In current redshift surveys, such measurements will constrain the lifetimes of low redshift quasars more accurately than measurements of the quasar autocorrelation function, because galaxies have much higher space densities than quasars. On scales smaller than 1 h−1 Mpc, the main contribution to ξQG comes from quasar/galaxy pairs in the same dark matter halo. The amplitude of ξQG depends both on the location of the host galaxy and on the density profile of other galaxies within the halo. As a result, measurements on these scales yield information about the processes responsible for fuelling supermassive black holes. At high redshifts our models predict that quasars of fixed luminosity are located in less massive halos than at low redshift. They are therefore less biased relative to galaxies of given luminosity or stellar mass. We have used the simulations to calculate the evolution of the quasar auto-correlation function. We find that models with quasar lifetimes in the range 106− 10 years provide a good match to the results of the 2dF QSO survey.