Multi-wavelength Studies of Gamma Ray Bursts (GRBs) and Associated Counterparts

Abstract

Gamma-ray bursts (GRBs) are brief but exceptionally bright flashes of γ-rays. They emit as much energy in a few seconds as thousands of Suns will produce in its entire lifetime. Even though they were discovered more than half a century ago, there are still unanswered questions about the nature of their central engines and progenitor systems. GRBs have traditionally been divided into short-duration and long-duration classes based on the T90 durations (the duration in which 90% of the fluence is collected) less than or greater than two, respectively. However, the T90 duration relies on multiple factors such as the sensitivity of the detector, variations in background levels, and the specific energy bands in which the burst is observed. Moreover, the proposed mechanisms involving massive star collapse for long GRBs and compact object mergers for short GRBs could not explain the recent observations. An unambiguous classification scheme correlated to progenitor models is still lacking. The overarching goal of this thesis is to solve the outstanding problem of GRB classification and uncover their progenitor diversity. A systematic multi-wavelength analysis of GRB properties from γ-rays to radio wavelengths is combined with advanced machine learning techniques to tackle this problem. This approach aims to identify correlations between observables that can reliably distinguish between intrinsic GRB subclasses. It also seeks to uncover potential subgroups within the traditionally defined duration categories. Through detailed case studies of ambiguous GRBs and multi-parameter analysis of GRB samples, the research presented in this thesis provides new insights into GRB progenitors. Unsupervised machine learning applied to light curves reveals distinct clusters that may correspond to different progenitor populations. The combination of comprehensive data analysis and sophisticated machine learning tools here represents a significant step forward in solving the long-standing classification problem and advancing our understanding of the progenitor system of GRBs. Looking ahead, future work will expand this research through inclusion of afterglow properties and larger GRB datasets from different instruments. The prospect of applying these techniques to upcoming wide-field surveys heralds new opportunities to definitively link GRB subclasses to progenitor systems. Moreover, the multimessenger observations of the bursts involving gravitational waves will further shed light on their progenitor systems.