A comprehensive analysis of fermi gamma-ray burst data. III. energy-dependent T90 distributions of GBM GRBs and instrumental selection effect on duration classification
Ying Qin,Enwei Liang,Yunfeng Liang,Shuangxi Yi,Lin Lin,Binbin Zhang,Jin Zhang,Houjun Lü,Ruijing Lu,LianZhong Lü,Bing Zhang
DOI: https://doi.org/10.1088/0004-637X/763/1/15
2013-01-01
Abstract:The durations (T-90) of 315 gamma-ray bursts (GRBs) detected with Fermi/GBM (8-1000 keV) up to 2011 September are calculated using the Bayesian Block method. We compare the T-90 distributions between this sample and those derived from previous/current GRB missions. We show that the T-90 distribution of this GRB sample is bimodal, with a statistical significance level comparable to those derived from the BeppoSAX/GRBM sample and the Swift/BAT sample, but lower than that derived from the CGRO/BATSE sample. The short-to-long GRB number ratio is also much lower than that derived from the BATSE sample, i.e., 1:6.5 versus 1:3. We measure T-90 in several bands, i.e., 8-15, 15-25, 25-50, 50-100, 100-350, and 350-1000 keV, to investigate the energy-dependence effect of the bimodal T-90 distribution. It is found that the bimodal feature is well observed in the 50-100 and 100-350 keV bands, but is only marginally acceptable in the 25-50 keV and 350-1000 keV bands. The hypothesis of bimodality is confidently rejected in the 8-15 and 15-25 keV bands. The T-90 distributions in these bands are roughly consistent with those observed by missions with similar energy bands. The parameter T-90 as a function of energy follows (T) over bar (90). E-0.20 +/- 0.02 for long GRBs. Considering the erratic X-ray and optical flares, the duration of a burst would be even longer for most GRBs. Our results, together with the observed extended emission of some short GRBs, indicate that the central engine activity timescale would be much longer than T-90 for both long and short GRBs and the observed bimodal T-90 distribution may be due to an instrumental selection effect.