Unveiling the origin of X-ray flares in Gamma-Ray Bursts

TL;DR

This study provides a detailed analysis of 113 X-ray flares in GRB afterglows, revealing their energy-dependent properties, temporal evolution, and spectral characteristics, and discusses their connection to prompt emission with no current comprehensive model.

Contribution

It introduces an extensive catalog and multi-band analysis of X-ray flares, highlighting their energy dependence, temporal evolution, and spectral features, advancing understanding of their origin.

Findings

Flares are narrower at higher energies following w~E^{-0.5}.

Flare width increases linearly with time, distinguishing them from prompt emission.

Peak luminosity decreases with time as L_{pk}~ t_{pk}^{-2.7}.

Abstract

We present an updated catalog of 113 X-ray flares detected by Swift in the ~33% of the X-ray afterglows of Gamma-Ray Bursts (GRB). 43 flares have a measured redshift. For the first time the analysis is performed in 4 different X-ray energy bands, allowing us to constrain the evolution of the flare temporal properties with energy. We find that flares are narrower at higher energies: their width follows a power-law relation w~E^{-0.5} reminiscent of the prompt emission. Flares are asymmetric structures, with a decay time which is twice the rise time on average. Both time scales linearly evolve with time, giving rise to a constant rise-to-decay ratio: this implies that both time scales are stretched by the same factor. As a consequence, the flare width linearly evolves with time to larger values: this is a key point that clearly distinguishes the flare from the GRB prompt emission. The flare 0.3-10 keV peak luminosity decreases with time, following a power-law behaviour with large scatter: L_{pk}~ t_{pk}^{-2.7}. When multiple flares are present, a global softening trend is established: each flare is on average softer than the previous one. The 0.3-10 keV isotropic energy distribution is a log-normal peaked at 10^{51} erg, with a possible excess at low energies. The flare average spectral energy distribution (SED) is found to be a power-law with spectral energy index beta~1.1. These results confirmed that the flares are tightly linked to the prompt emission. However, after considering various models we conclude that no model is currently able to account for the entire set of observations.