This paper looks for long-duration gamma-ray bursts (LGRBs) that do not show the bright supernova (SN) usually expected when a massive star dies. The authors used two decades of observations from the Neil Gehrels Swift Observatory to build a nearby sample (redshift z ≲ 0.35) where follow-up searches for supernovae and faint kilonovae are possible. Their goal was to decide whether SN-less LGRBs are rare oddities or a meaningful part of the burst population.

They compiled a volume-limited sample of 46 Swift bursts detected between 2005 and 2025. Using the usual 2-second dividing line for short versus long bursts, 35 were classed as long. The team split those into three groups: 10 with confirmed SN associations (standard collapsars), 5 with deep non-detections of any SN (examples include GRB 050724, 060505, 060614, 191019A, and 211211A), and 20 LGRBs with no SN reported but uncertain origins. The last group is the focus of their analysis.

To test each burst they compared high-energy properties, afterglow behavior, the local galaxy environment, and limits on accompanying SN or kilonova (KN) emission. Kilonovae are faint, short-lived optical/infrared flashes expected after some neutron star mergers. Some notable nearby long bursts (for example GRB 211211A and GRB 230307A) showed kilonova-like signals and occurred away from star-forming regions, strengthening the link to compact binary mergers rather than collapsars.

From their best (fiducial) sample the authors estimate a volumetric rate, uncorrected for beaming, of R = (0.5 ± 0.2) Gpc^-3 yr^-1. A gigaparsec cubed (Gpc^3) is a very large cosmic volume; this rate is comparable to the apparent rate of short GRBs and of ordinary LGRBs that have bright SNe with luminosities ≳ 10^50 erg s^-1. If the jets are collimated (beamed), as expected, correcting for that could raise the true rate and make it consistent with several compact binary merger channels.