Attributes of GRB Pulses: Bayesian Blocks Analysis of TTE Data; a Microburst in GRB 920229

Abstract

Bayesian Blocks is a new time series algorithm for detecting localized structures (spikes or shots), revealing pulse shapes, and generally characterizing intensity variations. It maps raw counting data into a maximum likelihood piecewise constant representation of the underlying signal. This bin-free method imposes no lower limit on measurable time scales. Applied to BATSE TTE data, it reveals the shortest know burst structure -- a spike superimposed on the main burst in GRB 920229 = Trigger 1453, with rise and decay timescales ~ few 100 microseconds.