Probing the Physics and Geometry of Active Galactic Nuclei

Abstract

Super Massive Black Holes (SMBH) inhabit the centre of every major galaxy, some of which power Active Galactic Nuclei (AGN) via mass accretion. In this thesis I explore two of the main approaches taken to study AGN, firstly the use of a wide sample, and secondly detailed observations of a single object. Multi-wavelength emission from AGN can be broken down into physical components which dominate in different wavebands, namely the dusty torus (IR), the accretion disc (optical/UV), the soft X-ray excess (UV/soft X-ray) and hot corona (hard X-rays). I first outline our new sample of around 700 SDSS Optical/UV/X-ray selected AGN, the SOUX sample, where the SDSS emission lines give single epoch black hole mass estimates while the UV/X-rays determine the continuum. I bin and stack the sample in mass and luminosity to look for trends as a function of these two important parameters. I perform detailed continuum fitting on the SOUX sample including the disc, soft X-ray excess and hot corona, and demonstrate clear issues in all current accretion flow models. Simple discs cannot match the UV extent of the most massive AGN irrespective of black hole spin as the high spin fits used in previous studies did not include the strong gravitational redshift which results from general relativistic ray tracing from the inner disc to the observer. I conclude that either there is a systematic over-prediction in black hole masses, or, more likely, the accretion flow in AGN does not take the form of a standard disc but may be fully Comptonised or take the form of something altogether different. I then switch gear to use the variable accretion flow emission to constrain the physical size scale of the dusty torus and Paschen Broad Line Region (BLR) from near-IR reverberation mapping. For Mrk 509, I develop a technique of photometric re-scaling and am able to place loose constraints on the size scales of these regions despite the challenging dataset. This technique is repeated with the much more robust data-set on Mrk 817, observed as part of the wider STORM2 collaboration. Both of these show that the Paschen BLR is co-spatial with the Balmer BLR, which is important as the Paschen BLR is now resolvable with GRAVITY data in a few objects, so can be used to test single epoch mass estimates. I then summarise the presented work and discuss potential avenues of investigation to take this further.