Analysis of Variability Phenomena in Soft X-Ray Sources

Abstract

In this thesis we present an analysis of X-ray variability phenomena in the soft X-ray emission from accreting black holes in Active Galactic Nuclei (AGN) and X-ray binaries (XRBs).

Following the discovery of Quasi-Periodic Eruptions we reconsider the variability in the low-mass AGN 2XMM J123103.2+110648. We apply the autocorrelation function to two observations and determine time-scales of ∼ 13. 52 and ∼ 14. 35 ks. Lightcurve modelling indicates a Gaussian model is preferable over a sinusoid, but the variability seen in 2XMM J123103.2+110648 cannot be definitively declared as a series of QPEs. Instead, this suggests a continuum of quasi-periodic variability from eruptions to oscillations. This opens the possibility of finding further sources that continue bridging this gap. Analysis of 47 observations of other low-mass AGN found no evidence of QPE- or QPO-like behaviour. We then apply machine learning to the discovery of new QPE sources. We search through archival data for as-yet undetected sources. Using a neural network and 14 time-domain features we classify lightcurves with accuracies of 94% with simulated data, and 98% on a sample of observational data. Analysis of 83,531 X-ray detections from the XMM Serendipitous Source Catalogue recovered lightcurves of known QPE sources and examples of variable stellar objects.

Then, we examined the effects of detector deadtime on the NICER XTI, and the evolution
of the XRB GX 339-4 as it transitioned through a soft flare. We found no significant effects
on the deadtime experienced by individual detectors and that even at the highest count rates considered the fractional deadtime is less than 1%. We examine features detected in power, frequency-lag, and energy-lag spectra across a series of 20 observations demonstrating the full range of countrates and hardness. We find systematic, qualitative changes in the timing properties associated with different phases of the flux and spectral evolution.

Date of Award	15 Feb 2024
Original language	English
Awarding Institution	University of Bristol
Supervisor	Andrew J Young (Supervisor) & Mark Birkinshaw (Supervisor)