Identification of patterns in cosmic-ray arrival directions using dynamic graph convolutional neural networks

We present a new approach for the identification of ultra-high energy cosmic rays from sources using dynamic graph convolutional neural networks. These networks are designed to handle sparsely arranged objects and to exploit their shortand long-range correlations. Our method searches for patterns in the arrival directions of cosmic rays, which are expected to result from coherent deflections in cosmic magnetic fields. The network discriminates astrophysical scenarios with source signatures from those with only isotropically distributed cosmic rays and allows for the identification of cosmic rays that belong to a deflection pattern. We use simulated astrophysical scenarios where the source density is the only free parameter to show how density limits can be derived. We apply this method to a public data set from the AGASA Observatory. (C) 2020 The Authors. Published by Elsevier B.V.

Automated summary

A new approach using dynamic graph convolutional neural networks is presented for identifying ultra-high energy cosmic rays from sources by searching for patterns in the arrival directions of cosmic rays. The method can discriminate between astrophysical scenarios with source signatures and those with isotropically distributed cosmic rays, allowing for the identification of cosmic rays belonging to a deflection pattern. The approach is demonstrated using simulated astrophysical scenarios with varying source densities to derive density limits.