Dark matter halo mass of AGN as a function of X-ray luminosity. Each panel corresponds to different redshift intervals, z=0-0.4, 0.4-0.9, 0.9-1.3. The data-points are a compilation of measurements of the mean dark matter halo mass of X-ray AGN from the literature. The grey shaded regions are the predictions of the GALFORM semi-analytic model for the formation of galaxies and supermassive black holes. The region demarcated by the black lines (higher normalisation curves) is for the version of GALFORM that includes two modes for growing black holes: one related to star-formation events (cold-gas accretion mode) and a second disjoint from star-formation that takes place in passive galaxies (hot-gas fueling mode). That model is conistent with the observations. For comparison also shown are the model predictions on the mean dark matter halos of AGN for the cold-gas accretion mode only (shaded regions demarcated with the red lines). A single accretion mode is not supported by the data. Both the cold and hot gas fueling modes of GALFORM are required to explain the observations. More details in Fanidakis, Georgakakis et al. (2013).

The large scale environment of AGN may provide important clues on the physical conditions under which supermassive black holes grow their masses. Studies of the environment of powerful broad-line QSOs for example, suggest typical dark matter halos, M_DM~10¹² (solar units). This is similar to the mass scale where baryons are turned into stars more efficiently in the Universe, thereby suggesting a relation between luminous QSOs, star-formation and the gas supply from large scales.

Contrary to broad line QSOs, moderate luminosity AGN selected at X-ray wavelengths are found to live, at least in an average sense, in more massive dark matter halos, M_DM~10¹³ (solar units). This mass scale corresponds to galaxy groups, suggesting that this type of environment may play an important role in the formation of black holes and the evolution of AGN.

There is also evidence that the clustering properties of AGN also depend on the level of obscuration of the central engine. This unexpected result questions the basic AGN Unification Model, which postulates an obscuring medium in the form of a torus, which is randomy oriented relative to the observer.

Further work is required to better constrain the above trends and interpret them in the context of physical models for the growth of supermassive black holes. Studies are also underway to place AGN in the context of galaxy formation and evolution. AGN may be transient events that occur in nearly all galaxies during their lifetime. In this case we can use our knowledge of the galaxy clustering and its dependence on e.g. stellar mass or star-formation rate, to interpret the large scale environment of AGN. This approach also requires knowledge of the properties (e.g. stellar mass, star-formation rate) of AGN host galaxies.