Bayesian Time-Resolved Spectroscopy of GRB Pulses

We performed time-resolved spectroscopy on a sample of 38 single pulses from 37 gamma-ray bursts detected by the Fermi/Gamma-ray Burst Monitor during its first 9 years of mission. For the first time a fully Bayesian approach is applied. A total of 577 spectra are obtained and their properties studied using two empirical photon models, namely the cutoff power law and Band model. We present the obtained parameter distributions, spectral evolution properties, and parameter relations. We also provide the result files containing this information for usage in further studies. It is found that the cutoff power law model is the preferred model, based on the deviance information criterion and the fact that it consistently provides constrained posterior density maps. In contrast to previous works, the high-energy power-law index of the Band model, $β$, has in general a lower value for the single pulses in this work. In particular, we investigate the individual spectrum in each pulse, that has the largest value of the low-energy spectral indexes, $α$. For these 38 spectra, we find that 60% of the $α$ values are larger than $-2/3$, and thus incompatible with synchrotron emission.Finally, we find that the parameter relations show a variety of behaviours. Most noteworthy is the fact that the relation between $α$ and the energy flux is similar for most of the pulses, independent of any evolution of the other parameters.