Maximum Likelihood Analysis of Clusters of Ultra-High Energy Cosmic Rays

Abstract

We present a numerical code designed to conduct a likelihood analysis for clusters of nucleons above 10**19 eV originating from discrete astrophysical sources such as powerful radio galaxies, gamma-ray bursts or topological defects. The code simulates the propagation of nucleons in a large-scale magnetic field and constructs the likelihood of a given observed event cluster as a function of the average time delay due to deflection in the magnetic field, the source activity time scale, the total fluence of the source, and the power law index of the particle injection spectrum. Other parameters such as the coherence length and the power spectrum of the magnetic field are also considered. We apply it to the three pairs of events above 4X10**19 eV recently reported by the Akeno Giant Air Shower Array (AGASA) experiment, assuming that these pairs were caused by nucleon primaries which originated from a common source. Although current data are too sparse to fully constrain each of the parameters considered, and/or to discriminate models of the origin of ultra-high energy cosmic rays, several tendencies are indicated. If the clustering suggested by AGASA is real, next generation experiments with their increased exposure should detect more than 10 particles per source over a few years and our method will put strong constraints on both the large-scale magnetic field parameters and the nature of these sources.