Y. S. Shirasaki

Chemical composition of primary cosmic rays with energies from 1015 to 1016.5 eV

Abstract The chemical composition of primary cosmic rays with energies from 1015 to 1016.5 eV, so called “knee” region, is examined. We have observed the time structures of air Cerenkov light associated with air showers at Mt. Chacaltaya, Bolivia, since 1995. The distribution of a parameter that characterizes the observed time structures is compared with that calculated with a Monte Carlo technique for various chemical compositions. Then the energy dependence of the average logarithmic mass numbers 〈 ln A 〉 of the primary cosmic rays is determined. The present result at 1015.3 eV is almost consistent with the result of JACEE (A≃12) and shows gradual increase in 〈 ln A 〉 as a function of the primary energy (A≃24 at 1016 eV). Form the comparison of the observational results with several theoretical models, we conclude that the supernova explosion of massive stars is a plausible candidate for the origin of cosmic rays around the “knee” region.

Monte Carlo simulation of air shower development for the study of cosmic ray composition

Abstract We have developed a Monte Carlo program to simulate the extensive air showers (EASs) in short computing time and to study the chemical composition of cosmic rays at the knee region. This code is particularly applied to the C erenkov experiment that measures the time structure of C erenkov light from the EAS. There are many EAS programs available for public use. However, they are not adequate for simulating large number of air shower, since they deal with the very complex interaction algorithm and consume a lot of time. Therefore much faster program should be developed adequately for our purpose. In this report, we discuss the interaction model used in our Monte Carlo code and make a comparison with the accelerator data and the widely used Monte Carlo code, CORSIKA. In addition, we propose a parameterization method that describes the fluctuation of air shower development. Using this method, we can calculate air shower development 106 times faster than the CORSIKA simulation code.