A Bayesian Approach to Black Hole Spin and Accretion Properties

Spin is an important parameter for black holes. It may be responsible for setting a limit for the inner edge of the accretion disc, producing relativistic jets and giving insights into black hole formation histories. For black holes in X-ray binaries, there is no consensus about their spins and significant disagreements over the existence of truncated discs during outburst. This thesis aims to address this debate, applying a fully Bayesian approach to modelling. Design goals were to have: self-consistency between model components; informative priors on column densities for interstellar medium absorption features based on XMM-Newton RGS data; informative priors on binary system parameters (mass, distance and inclination) from the best optical and radio measurements in the literature and relaxed assumptions on parameters where the physics is not well understood.

In Chapters 3 and 4, hard state XMM-Newton EPIC pn and NuSTAR observations from the 2018-2019 outburst of MAXI J1820+070 were used as a successful test case for the model. Markov Chain Monte Carlo and nested sampling results suggest a preference for low spins over very high (a > 0.9) spin, without the need for disc truncation.

In Chapter 5, the analysis was expanded to other systems (including hard and soft state observations). Column densities for 14 systems were obtained using XMMNewton RGS data. Three of these had good mass, distance and inclination measurements. In Chapter 6, their XMM-Newton EPIC pn spectra were analysed using the model. Results could only be obtained for GRO J1655-40, showing a strong preference for low spin. Later observations are consistent with retrograde spins but could be explained by a degree of disc truncation, obscuration of the inner disc or7 other changes at high accretion rates.

Conclusions and future work are presented in Chapter 7.