The Fermi satellite, launched in 2008, has two instruments: the Gamma-ray Burst Monitor (GBM, ) and the Large Area Telescope (LAT, ). The GBM measures the spectra of GRB in the energy range from 8 keV to 40 MeV, determining their position to ~ 5° accuracy. The LAT measures the spectra in the energy range from 20 MeV to 300 GeV, locating the source positions to an accuracy of < 1°. The GBM detects GRBs at a rate of ~ 250 per year, of which on average 20% are short bursts, while the LAT detects bursts at a rate of ~ 8 per year. The great strength of this combination is to provide the large field of view and high detection rate of the GBM extending to energies as low as the BAT in Swift, with the very high energy window of the LAT, which opens up a whole new vista into the previously almost unexplored GeV to sub-TeV range of GRBs.
Two unexpected features of the GeV emission of bursts were soon discovered by the Fermi-LAT. One is that the onset of the GeV emission is invariable delayed relative to the onset of the MeV emission (by a few seconds in LGRBs, and fractions of a second in SGRBs), e.g. [87, 88, 89, 90]. The other is that the GeV emission generally lasts for much longer then the MeV emission, decaying as a power law in time and lasting up to a 1000 s in some cases, i.e. well into the afterglow phase, including both LGRBs and SGRBs. The fact that GeV emission has been detected from a number of SGRBs is, in itself, also new. Remarkably, the GeV behavior of LGRBs and SGRBs is quite similar. This is not unexpected, since most of the GeV emission is produced in the afterglow phase, which is essentially a self-similar process. What is more unexpected is that the ratio of the total energy in the GeV range to MeV range is ~ 0.1 -0.5 for LGRBs, while it is 1 for SGRBs.
Bursts detected with the LAT have spanned a range of redshifts extending up to z = 4.3, with photon energies (in the burst rest frame) up to 10-130 GeV, the highest value so far being being that found for GRB 130427A , at a redshift z ~ 0.33. This is encouraging for the planned large Cherenkov Telescope Array (CTA) [93, 94], whose energy threshold may be as low as 50 GeV and whose detection rate of GRBs is estimated in the range 0.7-1.6 per year, based on the rate of Swift triggers. A roughly similar rate of detection is also expected for the High Altitude Water Cherenkov (HAWC) detector [95, 96], whose threshold is expected to be 10-20 GeV.