New insights on the duration distribution of long GRBs from Collapsars

According the Collapsar model long gamma-ray bursts (LGRBs) involve relativistic jets that puncture the envelope of a collapsing star, and produced the \gamma-rays after they break out. This model provides a theoretical framework for the well known association between LGRBs and massive stars. However although this association is supported by a wealth of observations, to this date there is no direct observational evidence for the emergence of the jet from the star. In other words there is no direct evidence for the Collapsar model. Here we show that a distinct signature of the Collapsar model is the appearance of a plateau in the duration distribution of the prompt GRB emission at times much shorter than the typical breakout time of the jet. This plateau is evident in the data of all major GRB satellites, and provides a direct evidence supporting the Collapsar model. It also enables us to place limits on the sizes and masses of LGRB progenitors; suggests the existence of a large population of choked (failed) GRBs; and indicates that the 2 s duration commonly used to separate Collapsars and non-Collapsars holds for BATSE and possibly Fermi GBM GRBs, but it is inconsistent with the duration distributions of Swift GRBs.