One possibility is to look for evidence of neutron irradiation in rock layers that were exposed at the time. The signal would be weak: 1) The gamma ray spectru

One possibility is to look for evidence of neutron irradiation in rock layers that were exposed at the time. The signal would be weak: 1) The gamma ray spectrum from GRBs looks like it has a long high-energy tail: http://www.mpa-garching.mpg.de/lectures/ADSEM/WS0304_Kienlin.pdf http://www.mpa-garching.mpg.de/lectures/ADSEM/WS0304_Kienlin... 2) For gamma energies about about 8 MeV, you start producing neutrons from gamma-nucleus interactions (this is why cancer patients who are treated with high-energy beams are very slightly radioactive for a while after each treatment: high energy photons blow neutrons out of a few of their nuclei, leaving short-lived daughter nuclei in their place.) 3) For an extinction-level gamma-ray event the GRB fluxes are estimated as ~10 kW/m^2 as the top of the atmosphere and 160 W/m^2 at the surface of the exposed side of the Earth. That's 160/1.6E-13 = 1E15 MeV/s.m^2. If 10E-6 gammas generate a neutron (totally ballpark guess, but the spectrum is only down by about 1E-3 from peak at around 10 MeV, and neutron production cross-sections at around 10 MeV are on the order of a milli-barn for nitrogen and some heavier elements, so a 1E-3 fraction going into neutron production is not unreasonable) that's 1E9 neutrons/m.s. Given GRBs last for a couple of seconds, therefore, one would expect to see a layer of neutron-activated rock that was exposed to a fair neutron flux. 4) But... all rock is exposed to a background flux of a few neutrons per second, so over a few hundred years there would be a background signal that swamped this burst signal. Ergo, it is unlikely that this mechanism could be used to detect a signature of GRB extinction events on Earth. The additional neutron activation signal would be completely swamped by the natural background signal in almost any reasonable circumstances.