The Evolution of AGN and Their Host Galaxies
Active galaxies have been in the forefront of astronomic research since their first discovery, at least 50 years ago (e.g. Schmidt, 1963; Matthews & Sandage, 1963). The putative supermassive black hole (SMBH) at their center characterizes their properties and regulates the evolution of these objects. In this thesis, I study the ‘demographics’ and ‘ecology’ of active galactic nuclei (AGN) in the context of their evolution and the interaction with their environments (mainly their host galaxy). The number density of AGN has been found to peak at 1 < z < 3 (e.g. Ueda et al., 2003; Hasinger et al., 2005; Richards et al., 2005; Aird et al., 2010), similar to the star formation history (e.g. Silverman et al., 2008a; Aird et al., 2010). However, when taking into account obscuration, faint AGN are found to peak at lower redshift (z ≤ 2) than that of bright AGN (z ≈ 2 − 3; e.g. Hasinger et al., 2005; Hopkins et al., 2007; Xue et al., 2011). This qualitative behaviour is also broadly seen in star-forming galaxies (e.g. Cowie et al., 1996) and is often referred to as ‘cosmic downsizing’, although this term has developed a number of usages with respect to galaxies (e.g. Bundy et al., 2006; Cimatti et al., 2006; Faber et al., 2007; Fontanot et al., 2009). Though this behaviour is well established up to z ≈ 3, the nature of how and when the initial seed of these AGNs were formed remains an open question. For this study, I use Chandra surveys to study some of the most distant AGN in the Universe (z > 3). The combination of two different size and depth Chandra surveys (Chandra- COSMOS and ChaMP) provides me with the largest to-date z > 3 AGN sample, over a wide range of rest-frame 2-10 keV luminosities [log (LX/erg s−1) = 43.3−46.0] and obscuration (NH = 1020 − 1023 cm−2). I find strong evidence about a strong decline in number density of X-ray AGN above z ≈ 3, and also the association of this decline with a luminositydependent density evolution (LDDE; e.g. Gilli et al., 2007). Especially at high redshifts, the different evolution models predict quite different numbers of AGNs. The large size and the wide X-ray luminosity range of this sample reduces the uncertainties of previous studies at similar redshifts making it possible to distinguish between the different models and suggest that observations appear to favour the LDDE model. The observed AGN downsizing behaviour seen via the measured X-ray luminosity function (XLF) could arise due to changes in the mass of the typical active SMBH and/or changes in the typical accretion rate. But how does the growth of SMBHs over cosmic time influence its environment? A powerful way to address this question is to compare the host galaxy properties over a wide range of AGN and accretion rate types. Radio-jets are one of the most prominent constituents of AGN as they can interact directly with the host galaxy. Although AGN with radio jets are rare (they make up to 10 per cent of the total AGN population) radio galaxies make up over 30 per cent of the massive galaxy population and it is likely that all massive galaxies go through a radio-loud phase, as the activity is expected to be cyclical (e.g Best et al., 2005). It is therefore, important to investigate the impact of radio jets on the host galaxy and particularly the star formation. The method I follow focuses on the comparison of the host galaxy properties between optically selected quasar samples, with and without strong radio emission associated with powerful radio-jets, matched in AGN luminosity. Herschel far-infrared observations are used to trace the star formation in the host galaxy, providing minimal AGN contamination. In my first approach, I have constructed a sample of radio-loud and radio-quiet quasars from the Faint Images Radio Sky at Twenty-one centimetres (FIRST) and the Sloan Digital Sky Survey Data Release 7 (SDSS DR7), over the H-ATLAS Phase 1 Area (9h, 12h and 14.5h). The main result of this work is that RLQs at lower AGN luminosities tend to have on average higher FIR and 250-μm luminosity with respect to RQQs matched in AGN luminosity and redshift. However, evolution effects could be strong as the quasars in this sample cover a wide range of redshifts (0.4 < z < 5). Therefore, I follow a second approach with the advantage of a QSO sample selection at a single redshift epoch, decomposing the evolution effects from the AGN/star-formation study. The results indicate that radio-jets in powerful QSOs can both suppress and enhance the star formation in their host galaxies. These fundings are consistent with a galaxy mass and jet-power dependence model. Then we expect more massive galaxies to have more star-formation for a given jet-power because their star-formation is more enhanced by the jet. Although radio-jets are the best candidates for a direct AGN impact to the host galaxy, many models refer to an AGN feedback associated with energetic AGN winds and outflows which are expected to suppress the star formation in powerful AGN when compared to the overall galaxy population. My results do not suggest star formation is suppressed in the hosts of optically selected QSOs at z ≈ 1, with more than 30 per cent of them being associated with strong star formation rates (SFR ≈ 350 M⊙ yr−1). Although different interpretations are possible, this result can be explained through periods of enhanced AGN activity and star-forming bursts, possibly through major mergers. However, optical QSOs comprise only a small fraction of the total AGN population. Even if the ‘unified model’ predicts that the host galaxy properties should not be affected by the viewing angle (type-1 vs. type-2 AGN), several studies have shown results supporting a scenario departing from the basic model. Investigating star formation in the hosts of 24 μm selected type-1&2 AGN, I found that the type-2 AGNs display on average higher star-formation rate than type-1 AGNs. This result is in agreement with previous studies suggesting an undergoing transition between a hidden growth phase and an unobscured AGN phase.