On Spectral Evolution and Temporal Binning in Gamma-Ray Bursts

TL;DR

This paper evaluates various temporal binning techniques for gamma-ray burst spectra, finding that Knuth and Bayesian blocks methods accurately reconstruct spectral evolution, while others have limitations, impacting the interpretation of GRB physics.

Contribution

It systematically compares multiple binning methods for GRB spectral analysis, identifying the most effective techniques for capturing intrinsic spectral evolution.

Findings

Knuth and Bayesian blocks methods accurately recover spectral evolution

Signal-to-noise binning often fails to represent true spectral changes

Integrated pulse properties may not reflect physical evolution without detailed spectral analysis

Abstract

The understanding of the prompt $\gamma$-ray spectra of gamma-ray bursts (GRBs) is of great importance to correctly interpreting the physical mechanisms that produce the underlying event as well as the structure of the relativistic jet from which the emission emanates. Time-resolved analysis of these spectra is the main method of extracting information from the data. In this work, several techniques for temporal binning of GRB spectra are examined to understand the systematics associated with each with the goal of finding the best method(s) to bin lightcurves for analysis. The following binning methods are examined: constant cadence (CC), Bayesian blocks (BB), signal-to-noise (S/N) and Knuth bins (KB). I find that both the KB and BB methods reconstruct the intrinsic spectral evolution accurately while the S/N method fails in most cases. The CC method is accurate when the cadence is not too coarse but does not necessarily bin the data based on the true source variability. Additionally, the integrated pulse properties are investigated and compared to the time-resolved properties. If intrinsic spectral evolution is present then the integrated properties are not useful in identifying physical and cosmological properties of GRBs without knowing the physical emission mechanism and its evolution.