2017ApJ...846..152V

A gamma-ray burst (GRB) jet running into an external medium is expected to generate luminous GeV-TeV emission lasting from minutes to several hours. The high-energy emission results from inverse Compton upscattering of prompt and afterglow photons by shock-heated thermal plasma. At its peak the high-energy radiation carries a significant fraction of the power dissipated at the forward shock. We discuss in detail the expected TeV luminosity, using a robust "minimal" emission model. Then, using the statistical properties of the GRB population (luminosity function, redshift distribution, afterglow energy), we simulate the expected detection rates of GRBs by current and upcoming atmospheric Cherenkov instruments. We find that GRBs exploding into a low-density interstellar medium must produce TeV emission that would have already been detected by the currently operating Cherenkov telescopes. The absence of detections is consistent with explosions into a dense wind of the GRB progenitor. If, as suggested by the recent analysis of Fermi LAT data, the typical environment of long GRBs is a Wolf-Rayet progenitor wind with the density parameter A∼10¹¹ g cm^–1, then 10%-20% of the bursts that trigger the space-borne detectors should also be detectable by the upcoming Cherenkov Telescope Array (CTA) under favorable observing conditions. Since absorption by the extragalactic background light limits the detectability above 0.1 TeV for all but the most nearby bursts (z 1), the reduced energy threshold of CTA is the key improvement over current instruments, which should increase the number of detectable bursts by at least a factor of 3 compared with currently operating facilities.