Classifying groups of PCR nuclei with energies of 1015–1016 eV according to the spatial-angular distribution of EAS Cherenkov light

Abstract

Multidimensional Bayesian classification criteria are proposed for groups of nuclei of primary cosmic rays based on characteristics of spatial-angular distribution of Cherenkov light of extensive air showers. These criteria allow us in principle to separate no less than 50–60% of the primary protons from heavier nuclei; the classification errors for primary protons and groups of nitrogen and iron nuclei total no more than several per cent. New parameters that substantially improve the separability of classes of showers are also found for angular images of Cherenkov light.