Coupled black hole and galaxy formation in the young universe

Abstract

This thesis presents results from three major imaging studies designed to determine the properties of massive galaxies selected at high redshift. The most massive known galaxies at high redshift (i.e. radio galaxies), the most massive known starbursts at high redshift (i.e. submillimetre galaxies), and the most massive known black holes at high redshifts (i.e. quasars) have been targeted, with the aim of better establishing the relationship between these different classes of massive object.The results from deep K—band imaging of the most luminous radio galaxies at z ~ 2 and the brightest submillimetre sources in the 8- mjy survey are presented. Morphological properties are measured and compared to explore whether submillimetre hosts are the progenitors of today's most massive ellipticals. Two-dimensional modelling finds that the high-redshift radio galaxies have surface brightness distributions consistent with those expected for classic ellipticals (Sersic parameter, n = 4), with n in the range 2.5 < n < 5.75 and < n >= 4.04±0.27. In contrast, submillimetre galaxies are found to be disks (Sersic parameter, n = 1) with n in the range 1.0 < n < 2.5 and <n>=1.46±0.16. The half-light radii of the radio hosts are distributed over a range 2 < r1/2 < 16kpc with < r1/2 >= 7.85 ± 1.2 kpc, while submillimetre sources are smaller, with 1 < rq/2 < 6 kpc and < rq/2 >= 3.15 ± 0.29 kpc. The z ~ 2 radio galaxies are found to be, on average, a factor ~ 2 smaller than massive radio galaxies at z < 1 and follow a Kormendy (/ie — re) relation similar to that of 3C-type galaxies at low redshift after evolutionary corrections, assuming a burst-formation model at z = 3, consistent with the median redshift of the submillimetre sources. The luminosities of the submillimetre galaxies are ~ 1.5 magnitudes brighter than Lyman-break galaxies at comparable redshifts, indicating that submillimetre galaxies are significantly more massive. Finally, surface mass densities for both source types are found to be consistent with quiescent elliptical galaxies at high redshift, and not star-forming or Lyman-break galaxies.The results are presented of a study that uses the 3CRR sample of radio-loud ac¬ tive galactic nuclei to investigate the evolution of the black-hole:spheroid mass ratio in the most massive early-type galaxies from 0 < z < 2. Radio-loud unification is ex¬ ploited to obtain virial (linewidth) black hole mass estimates from the 3CRR quasars, and stellar mass estimates from the 3CRR radio galaxies, thereby providing black hole and stellar mass estimates for a single population of early-type galaxies. At low redshift (z<l) the 3CRR sample is consistent with a black-hole:spheroid mass ratio of Mbh/Mbulge ~ 0.002, in good agreement with that observed locally for quiescent galaxies of similar stellar mass (Mbuige — 5 x 10u M0). However, over the redshift interval 0 < z < 2 the 3CRR black-hole:spheroid mass ratio is found to evolve as Mbh/Mbuige oc (1 + z)2-07±0-76/ reaching Mbh/Mbulge ~ 0.008 by redshift z ~ 2. If con¬ firmed, the detection of evolution in the 3CRR black-hole:spheroid mass ratio further strengthens the evidence that, at least for massive early-type galaxies, the growth of the central supermassive black hole may be completed before that of the host spheroid.The results from an ongoing project to constrain the evolution of the blackhole:spheroid mass ratio at z — 3,4 are presented. At these redshifts, quasars and their hosts provide a unique window on this evolution, as the only class of object in which both galaxy and black hole masses can be directly measured. Additionally, the black-hole:bulge mass ratios of flat spectrum quasars at z ~ 1.5 are used to deter¬ mine whether the BLR in AGN possesses a spherical or disk-like morphology. It is demonstrated that both black hole and host galaxy masses can be obtained from emis¬ sion linewidths and deep high-quality A-band imaging respectively. These pilot data represent the first attempt to constrain the black-hole:spheroid mass ratio as part of a statistically significant sample at z = 3,4, a crucial epoch of galaxy formation and evolution.