Tonima Tasnim Ananna

A new X-ray population synthesis model, its physical implications, along with detailed analysis of AGN X-ray spectral parameter space are presented in this work. This population synthesis model uniquely fits all the newest observed constraints from high energy X-ray bands such as NuSTAR and Swift-BAT, and predicts a Compton-thick fraction of 50−56% in the local Universe.

The new X-ray population synthesis model was computed using a neural network. Given an input distribution of AGN spectral parameters, this neural network converges onAGN populations that can fit the cosmic X-ray background (CXB). Other observed quantities such as AGN number counts and Compton-thick fractions are then used for validation to choose the correct solution from all possible matches to the X-ray background.

A measure of the average radiative efficiency and AGN ionization contribution based on the new population synthesis model are also presented in this work. We find somewhat higher values of radiative efficiency with our model than was estimated by previous models, due in part to higher space densities of obscured objects in the new model, as well as to using a newer (lower) reported value for the local black hole mass density. Recent results suggest that AGN in late stages of galaxy mergers tend to be heavily obscured. Therefore more obscured objects may spin up as a result of recent mergers, leading to higher radiative efficiencies.

We find that some regions of the X-ray spectral parameter space can produce accept-able fits to the CXB, while others will overproduce or underproduce parts of the CXB independent of the underlying AGN population. A full exploration of acceptable spectral parameter spaces is done using a Bayesian technique. We identify regions of the AGN spectral parameter space that never produce a fit to the CXB and thus can be rejected. These findings are compared to observed AGN spectral parameter distributions in surveys.

This work also presents a new multiwavelength study of AGN spectra in order to calculate photometric redshifts for Stripe 82X, one of the largest volume X-ray surveys to date. The calculation of these photometric redshifts informs us on the population types found in wide-shallow fields, which is necessary to understand the populations accessed in the wider layers of wedding cake surveys, and the techniques and AGN templates used are applicable to future wide-area surveys by LSST and eROSITA.