Detection of afterglow emission up to 100 GeV through a stacking analysis of gamma-ray bursts

Abstract

High-energy gamma-ray (>GeV) emission of gamma-ray bursts (GRBs) is very important in probing the jet evolution and particle acceleration of GRBs. The observations of high-energy photons are limited except for a few very bright GRBs, hindering precise measurements of the spectral and temporal evolutions of GRBs. Here we report the detection of high-energy gamma-ray emission up to 100 GeV with Fermi-LAT using a stacking analysis of a collection of 330 GRBs. High significance detection of the emission has been found, and the precise light curves and energy spectra can be measured. The light curves and time-resolved spectra of the sub-sample of 220 LAT individually detected GRBs can be well explained by the standard afterglow emission from a population of GRBs with both synchrotron and synchrotron self-Compton mechanisms, assuming a distribution of initial Lorentz factors. However, the emission of the relatively weak sample of the 110 LAT individually undetected GRBs cannot be well reproduced in the same framework, indicating the existence of possible energy injection effect in the GeV band for the first time. The observations hence provide new insights in understanding the high-energy emission of GRBs.

Figures (10)

• Figure 1: The $10^{\circ} \times 10^{\circ}$ significance maps of the stacked GRBs smoothed with a $0.4^{\circ}$ Gaussian kernel. For each GRB, photons in the region of the same size centered on its trigger position within a time duration from the trigger time $T_0$ to $T_0 + 50,000$ s are collected. Panel a is for the 220 LAT detected GRBs with photon energies from 0.1 to 100 GeV, and panel b is for the 110 LAT undetected GRBs with energies from 1 to 10 GeV.
• Figure 2: SEDs at different time for the 220 LAT detected GRBs. Five time intervals are adopted: $T_0+[0,10]$ s, $T_0+[10,100]$ s, $T_0+[100,1000]$ s, $T_0+[1000,10000]$s, and $T_0+[10000,50000]$ s. Panel (a) shows the results obtained for all photons, and panel (b) shows the results after removing photons within $T_{90}$ of each GRB. Dashed lines and shaded bands show the best-fitting results and the $\pm1\sigma$ uncertainty bands with an ECPL model.
• Figure 3: Light curves of the stacked GRBs and the corresponding fitting results. The light curves for all photons and for those with $T_{90}$ photons subtracted are shown in panels a and b for the 220 LAT detected GRBs, and in panels c and d for the 110 LAT undetected GRBs, respectively. All light curves are fitted using a multi-piece power-law model, with fitting parameters being given in Table \ref{['tab:lc']}.
• Figure 4: Stacked SEDs and light curves from GRB population with log-normal initial Lorentz factor distribution. Panels a, b and c present the stacked model SEDs and light curves for the high initial Lorentz factor ($\Gamma_0 \sim 130-800$) populations, corresponding to the 220 LAT detected GRBs. Panels d and e show the corresponding stacked model SEDs and light curves for the low initial Lorentz factor ($\Gamma_0 \sim 60-250$) populations, which are derived for the 110 LAT undetected GRBs.
• Figure A1: The distributions of TS values of pseudo-experiments, for the 220 LAT detected sample (panel a) and the 110 LAT undetected sample (panel b). The solid line in each panel is the best-fitting $\chi^2_{\nu}$ distribution with degree-of-freedom $\nu$.