Classification of Photospheric Emission in Short GRBs

Hüsne Dereli-Bégué, Asaf Pe'er, Felix Ryde

Published 2020-02-15Version 1

In order to better understand the physical origin of short duration gamma-ray bursts (GRBs), we perform time-resolved spectral analysis on a sample of 70 pulses in 68 short GRBs with burst duration $T_{90}\lesssim2$~s detected by the \textit{Fermi}/GBM during the first 11 years of its mission. We apply Bayesian analysis to all spectra that have statistical significance $S\ge15$ within each pulse and apply a cut-off power law (CPL) model. For our analysis, we identify the maximal values of the low energy spectral index, $\alpha_{\rm max}$. Under the assumption that the main emission mechanism is the same throughout each pulse, such an analysis is indicative of pulse emission. We find that $\sim$1/3 of short GRBs are consistent with a pure, non-dissipative photospheric model, at least, around the peak of the pulse. This fraction is larger compare to the corresponding one (1/4) obtained for long GRBs. For those bursts which are compatible with a pure photospheric origin, we find (i) a bi-modal distribution in the values of the Lorentz factors and the hardness ratios; (ii) an anti-correlation between $T_{90}$ and the peak energy, $E_{\rm pk}$: $T_{90} \propto E_{\rm pk}^{-0.35\pm0.01}$. This correlation disappears when we consider the entire short GRB sample. Our results thus imply that short GRB population may in fact be composed of two separate populations: a continuation of the long GRB population to shorter duration, and a separate population with distinct physical properties. Furthermore, thermal emission is initially ubiquitous, but is accompanied at longer times by additional radiation (likely synchrotron), making time-resolved spectroscopy absolutely necessary in analyzing GRB pulses.