The Environments of Active Galactic Nuclei

Near-infrared and optical CCD images of radio-loud quasars, radio-quiet quasars and radio galaxies (RLQs, RQQs, and RGs respectively) are presented. The quasar samples are chosen such that they are well matched in the V-z plane, whilst the RGs are selected so that their radio properties (5 GHz radio power and spectral index)are matched to those of the RLQs. The images of all three samples have been analysed using detailed 2-dimensional modelling to extract the parameters of the host galaxies. The low nuclear:host ratios displayed by quasars at 2.2gm combined with the consistent depth of the images have enabled the reliable determination of the luminosities, scale lengths and even morphological type of the majority of the host galaxies. In fact tests of the repeatability of the galaxy model fitting with increasing nuclear:host ratio indicate that, with ground-based seeing, it is only at near-infrared wavelengths that the properties of quasar host galaxies can be determined with acceptable reliability, although photometric information has been extracted from some of the optical images. The host galaxies of the RLQs and RGs are found to be almost universally bulge dominated systems, being best fitted by an r1/4 law surface brightness distribution thus lending clear support for the unification of these two classes of AGN via orientation. The radio-quiet quasars are found in a mixture of disc and bulge dominated host galaxies. Three out of the four RQQs in this study found to have bulge dominated hosts are known to contain weak radio sources suggesting that such RQQs may be capable of becoming powerful radio sources. Comparison of the luminosities of the host galaxies of the RLQs, RQQs and RGs reveals a picture of relative uniformity, in contrast to several previous optical studies. Generally the host galaxies are larger and more luminous than ordinary galaxies. The K-band luminosities of the host galaxies are between 2 and 3 times more luminous than an M galaxy indicating that they are drawn from the exponential tail of the K-band galaxy luminosity function. No significant difference is found between the luminosity distributions of the RLQ, RQQ and RG host galaxies. All the host galaxies have a half-light radius r112 > 10 kpc. In addition the host galaxies of all three classes of AGN display a - r112 relation identical in both slope and normalization to that displayed by brightest cluster galaxies. The host galaxy colours are entirely consistent with those predicted for a passively evolving 13.5 Gyr old elliptical galaxy. They display less scatter than those presented in Kotilainen & Ward (1994). Also none of the host galaxies are redder than the reddest colours predicted by the models of Rocca—Volmerange & Guiderdoni (1987) unlike those of Kotilainen & Ward (1994. The host galaxies are good standard candles at K displaying a K-z relation essentially identical to that found for powerful radio galaxies, but fainter by ~ 0.5 mag. The 2—D modelling of the RGs has shown that at K there is a significant contribution from an unresolved nuclear component in all cases. This nuclear contribution accounts for the host—galaxy K—z relation being fainter at low redshifts than that deduced from raw aperture photometry. The B—K colour of the nuclear emission is consistent with it arising from a dust reddened quasar. This result together with the morphological information strongly supports the 'unified scheme' as proposed by Barthel (1989) and that the RGs represent the unbeamed parent population of the RLQs. Further work based upon the existing data is suggested and observations to test the theory concerning the nature of the nuclear emission seen at K in the RGs are outlined