Abstract
Active galactic nuclei (AGN) are the extremely luminous hearts of a subset of galaxies we call active galaxies. It is commonly accepted that they contain supermassive black holes (SMBHs) at their centres, surrounded by a luminous accretion disk and extended gas regions at various radii from the black hole. One such region is the broad-line region (BLR), which contains high velocity clouds that produce the Doppler shifted emission lines. In this Thesis we present new developments in the field of studying AGN with a method called reverberation mapping. Reverberation mapping compares the variations in the accretion disk flux with the variations in the line flux of certain lines of the BLR, allowing to find an AGN’s BLR size; the distance between a SMBH and its BLR.
We simulate light curves for the central regions of AGN based on measurements we expect to obtain from the photometric Dark Energy Survey (DES), and the Australian spectroscopic counterpart, OzDES, to predict how well we can perform the reverberation mapping analysis using this dataset. We use a Markov Chain Monte Carlo (MCMC) based software (JAVELIN; Zu, Kochanek & Peterson, 2011) to obtain estimates for these distances, and test reliability when the emission line responds to changes in the accretion disk flux in a different way than that expected by the program. We find that JAVELIN reliably recovers the BLR size regardless of the inferred emission line response. This is important, as the true emission line response is unknown.
It has been observed that AGN with similar luminosities tend to recover similar BLR sizes, which is consistent with theoretical predictions. We take advantage of this behaviour, and use the output from Javelin in a Bayesian stacking analysis in order to obtain BLR size estimates for groups of AGN whose individual BLR size recoveries are not reliable. We find that the Bayesian stacking method is a more precise and accurate way of estimating the BLR size when compared to BLR size estimates from individual objects. In order to apply this method to the real survey and optimise the binning of the AGN into groups of similar luminosity, we simulated AGN with redshift and absolute magnitude distributions consistent with the DES/OzDES sample. We then apply the stacking analysis on simulated Mg ii light curves to find the optimal grouping of the sample. We find that approximately 95% of the AGN where Mg ii can be used for reverberation mapping can contribute to obtaining time lag estimates for AGN when using stacking.
Veileder: Førsteamanuensis Mark Dijkstra, Institutt for teoretisk astrofysikk, UiO
Medveileder: Postdoktor Signe Riemer-Sørensen, Institutt for teroretisk astrofysikk, UiO & Tamara Davis, University of Queensland, Australia
Intern sensor: Professor David F. Mota, Institutt for teoretisk astrofysikk, UiO
Ekstern sensor: Associate Professor Marianne Vestergaard, Niels Bohr Institute