Natural limit on the gamma/hadron separation for a stand alone air Cherenkov telescope

Abstract

The gamma/hadron separation in the imaging air Cherenkov telescope technique is based on differences between images of a hadronic shower and a gamma induced electromagnetic cascade. One may expect for a large telescope that a detection of hadronic events containing Cherenkov light from one gamma subcascade only is possible. In fact, simulations show that for the MAGIC telescope their fraction in the total protonic background is about 1.5% to 5.2% depending on the trigger threshold. It has been found that such images have small sizes (mainly below 400 photoelectrons) which correspond to the low energy primary gamma's (below 100 GeV). It is shown that parameters describing shapes of images from one subcascade have similar distributions to primary gamma events, so those parameters are not efficient in all methods of gamma selection. Similar studies based on MC simulations are presented also for the images from 2 gamma subcascades which are products of the same pi^0 decay. The ratio of the number of the expected background from false gamma and one pi^0 to the number of the triggered high energy photons from the Crab direction has been estimated for images with a small alpha parameter to show that the occurrence of this type of protonic shower is the reason for the difficulties with true gamma selection at low energies.