Supermassive Black Hole Activity in the Cosmic Evolution Survey

Abstract

I investigate active supermassive black holes, also called active galactic nuclei (AGNs). My tool for this work is the Cosmic Evolution Survey (COSMOS), a deep multiwavelength survey over 2 deg² of the sky. I describe the COSMOS AGN optical spectroscopy campaign, and present the largest AGN sample to date with full multiwavelength (radio, IR, optical, UV and X-ray) spectral energy distributions. Studying the COSMOS AGN sample reveals a unified model for supermassive black hole activity based on accretion rate, as shown by the following main results. (1) Classically “obscured” (Type 2) AGNs are more prevalent at higher redshifts and lower luminosities, suggesting that these objects accrete through low-level stochastic disk feeding by their hosts. (2) The prescence of broad emission lines in an AGN requires a minimum accretion rate (L/L(Edd) > 0.01). Broad-line (Type 1) AGNs in COSMOS span a large range of accretion rates (0.01 < L/L(Edd) < 1), in contrast to results from previous, shallower surveys, and broad-line AGNs become more optically luminous as accretion rate increases. (3) Lineless, “optically dull” AGNs have very different SEDs than broad-line and narrow-line AGNs, with comparatively brighter X-ray emission, redder optical continua, no infrared hot dust, and stronger radio emission. While up to 2/3 of optically dull AGNs may be “normal” AGNs diluted by extranuclear host galaxy light, at least 1/3 are best described as unobscured, intrinsically weak AGNs. (4) At low accretion rates, material accreting onto an AGN changes from a thin disk to an advection-dominated flow near the black hole, resulting in very different observed properties: the broad-line region disappears, radio jets become more important, and the hot dust signature changes. In contrast to previous unification models, observations indicate that most of the narrow-line and lineless AGNs in COSMOS are best described as weakly accreting AGNs. We conclude by noting a few predictions and observational tests to further investigate our model of AGN unification by accretion rate.