The galactic black hole low-mass x-ray binaries: an observational study

Abstract

Black hole low-mass X-ray binaries (BH-LMXBs) are pivotal cosmic laboratories for understanding accretion processes around black holes and the life cycles of black holes. However, our comprehension of these systems, including their fundamental properties and their distribution within our Galaxy, is often complicated by uncertain distance measurements and significant observational biases. A comprehensive understanding of their formation and evolution requires precise measurements of observational parameters, such as distance and accretion disk inclination, which are often poorly constrained. This dissertation confronts these challenges through a multi-faceted observational study. A robust statistical framework is developed to derive dependable distances by modeling X-ray spectra of BH-LMXB outbursts, incorporating general relativistic corrections and systematic uncertainty propagation. This method is applied to a large sample of sources using archival data from multiple X-ray missions. To address inherent observational selection effects and systematic biases, an extensive suite of simulations was conducted and a bias corrector was found. Lastly, this dissertation explores a novel approach to determine the accretion disk inclination by applying machine learning models to the X-ray light curves of these systems, the method entailed physics-inspired data augmentation and advanced imputation techniques in the hope of overcoming limitations of sparse datasets. The application of the new distance estimation framework produced a catalog of reliable distances for 26 BH-LMXBs. The subsequent bias-corrected analysis of their Galactic distribution provides a refined view of the population, revealing structures that trace the Galaxy spiral arms and hinting at a hidden population of sources near the Galactic plane. The resulting distribution of sources in relation to the Galactic plane is most consistent with a hybrid model for black hole formation, where some systems receive high natal kicks and others receive very small kicks. The machine learning investigation, while highlighting current data limitations, allowed us to lay the groundwork for future advancements. Collectively, these studies provide improved tools for the astronomical community, offer a refined understanding of BH-LMXB demographics, and contribute new insights into black hole formation mechanisms. Moreover, this work lays the foundation for continued exploration of these fascinating systems with upcoming multi-messenger facilities and evolving computational methods.