The pulse luminosity function of Swift gamma-ray bursts
The complete Swift Burst Alert Telescope and X-Ray Telescope light curves of 118 gamma-ray bursts (GRBs) with known redshifts were fitted using the physical model of GRB pulses by Willingale et al. to produce a total of 607 pulses. We compute the pulse luminosity function utilizing three GRB formation rate models: a progenitor that traces the cosmic star formation rate density (CSFRD) with either a single population of GRBs, coupled to various evolutionary parameters, or a bimodal population of high- and low-luminosity GRBs, and a direct fit to the GRB formation rate excluding any a priori assumptions. We find that a single population of GRB pulses with an evolving luminosity function is preferred over all other univariate evolving GRB models, or bimodal luminosity functions in reproducing the observed GRB pulse L-z distribution and that the magnitude of the evolution in brightness is consistent with studies that utilize only the brightest GRB pulses. We determine that the appearance of a GRB formation rate density evolution component is an artifact of poor parametrization of the CSFRD at high redshifts rather than indicating evolution in the formation rate of early epoch GRBs. We conclude that the single brightest region of a GRB light curve holds no special property, by incorporating pulse data from the totality of GRB emission we boost the GRB population statistics by a factor of 5, rule out some models utilized to explain deficiencies in GRB formation rate modelling, and constrain more tightly some of the observed parameters of GRB behaviour.