Distributions of Peak Flux and Duration for Gamma-Ray Bursts

Abstract

Many of the important conclusions about Gamma-Ray Bursts follow from the distributions of various quantities such as peak flux or duration. We show that for astrophysical transients such as bursts, multiple selection thresholds can lead to various forms of data truncation, which can strongly affect the distributions obtained from the data if not accounted for properly. Thus the data should be considered to form a multivariate distribution. We also caution that if the variables forming the multivariate distribution are not statistically independent of each other, further biases can result. A general method is described to properly account for these effects, and as a specific example we extract the distributions of flux and duration from the BATSE 3B Gamma-Ray Burst data. It is shown that properly accounting for the aforementioned biases tends to increase the slope of the $\log{N}$-$\log{S}$ relation at low values of $S$, and dramatically increases the number of short duration bursts.