STRATEGIES FOR SELECTION OF SMALL GAMMA-RAY SHOWERS USING THE IMAGING-ATMOSPHERIC-CHERENKOV-TECHNIQUE

Methods of selecting small candidate gamma-ray events obtained using t he Imaging Atmospheric Cherenkov Technique are presented. Three approa ches are described: a variant of the standard 'Supercuts' analysis tec hnique, a parameter weights approach, and a neural network approach. A cut on the parameter combination Length/Size is developed and shown t o be effective at rejecting background due to muons. The techniques ar e applied to observations of Markarian 421 taken during 1995 and 1996 with the Whipple 10 m reflector. The efficiency and stability of the t hree approaches are compared, and the Supercuts variant is shown to be the most robust. A signal is detected in the energy range 170 GeV to 300 GeV at a significance level of 6.5 sigma, in 7 hours of observatio n (the standard analysis gives 15.2 sigma for E > 300 GeV). This repre sents the first significant detection of a source at an energy thresho ld below 200 GeV using the Whipple 10 m system. (C) 1997 Elsevier Scie nce B.V.