Gamma-ray burst minimum variability timescales with Fermi /GBM

Gamma-ray bursts (GRBs) have traditionally been classified by duration as long (LGRBs) or short (SGRBs), with the former believed to originate from massive star collapses and the latter from compact binary mergers. However, events such as the SGRB 200826A (coming from a collapsar) and the LGRBs 211211A and 230307A (associated with a merger) suggest that duration-based classification could sometimes be misleading. Recently, the minimum variability timescale (MVT) has emerged as a key metric for classifying GRBs. We calculated the MVT, defined as the full width at half maximum (FWHM) of the narrowest pulse in the light curve, using an independent dataset from Fermi /GBM, and we compared our results with other MVT definitions. We updated the MVT-T_ plane and analysed peculiar events such as long-duration merger candidates 211211A, 230307A, and other short GRBs with extended emission (SEE-GRBs). We also examined extragalactic magnetar giant flares (MGFs) and explored possible new correlations with peak energy. We used the mepsa algorithm to identify the shortest pulse in each GRB light curve and measured its FWHM. We calculated the MVT for around 3700 GRBs, 177 of which have spectroscopically known redshift. The SEE-GRBs and SGRBs share similar MVTs (from a few tens of to a few hundred milliseconds, indicating a common progenitor, while extragalactic MGFs exhibit even shorter values (from a few milliseconds to a few tens of milliseconds). Our MVT estimation method consistently yields higher values than another existing technique, the latter aligning with the pulse rise time. For LGRBs, we confirm the correlations of MVT with peak luminosity and Lorentz factor. We confirm that although MVT alone cannot determine the GRB progenitor, it is a valuable tool when combined with other indicators, as it helps flag long-duration mergers and distinguish MGFs from typical SGRBs.