A method for determining AGN accretion phase in field galaxies

Recent observations of active galactic nucleus (AGN) activity in massive galaxies (log M*/M⊙ > 10.4) show the following: (1) at z < 1, AGN-hosting galaxies do not show enhanced merger signatures compared with normal galaxies, (2) also at z < 1, most AGNs are hosted by quiescent galaxies and (3) at z > 1, the percentage of AGNs in star-forming galaxies increases and becomes comparable to the AGN percentage in quiescent galaxies at z∼2. How can major mergers explain AGN activity in massive quiescent galaxies that have no merger features and no star formation to indicate a recent galaxy merger? By matching merger events in a cosmological N-body simulation to the observed AGN incidence probability in the COSMOS survey, we show that major merger-triggered AGN activity is consistent with the observations. By distinguishing between 'peak' AGNs (recently merger-triggered and hosted by star-forming galaxies) and 'faded' AGNs (merger-triggered a long time ago and now residing in quiescent galaxies), we show that the AGN occupation fraction in star-forming and quiescent galaxies simply follows the evolution of the galaxy merger rate. Since the galaxy merger rate drops dramatically at z < 1, the only AGNs left to be observed are the ones triggered by old mergers that are now in the declining phase of their nuclear activity, hosted by quiescent galaxies. As we go towards higher redshifts, the galaxy merger rate increases and the percentages of 'peak' AGNs and 'faded' AGNs become comparable.