Identifying nearby sources of ultra-high-energy cosmic rays with deep learning

TL;DR

This paper introduces a deep learning approach using convolutional neural networks to analyze the arrival directions of ultra-high-energy cosmic rays, significantly improving detection sensitivity for nearby sources compared to traditional methods.

Contribution

The authors develop a neural network-based method for identifying cosmic ray sources, outperforming angular power spectrum techniques and demonstrating robustness across various models and scenarios.

Findings

Deep learning method enhances detection sensitivity.

Neural networks outperform traditional angular power spectrum analysis.

Robust performance across different magnetic field models and compositions.

Abstract

We present a method to analyse arrival directions of ultra-high-energy cosmic rays (UHECRs) using a classifier defined by a deep convolutional neural network trained on a HEALPix grid. To illustrate a high effectiveness of the method, we employ it to estimate prospects of detecting a large-scale anisotropy of UHECRs induced by a nearby source with an (orbital) detector having a uniform exposure of the celestial sphere and compare the results with our earlier calculations based on the angular power spectrum. A minimal model for extragalactic cosmic rays and neutrinos by Kachelrie{\ss}, Kalashev, Ostapchenko and Semikoz (2017) is assumed for definiteness and nearby active galactic nuclei Centaurus A, M82, NGC 253, M87 and Fornax A are considered as possible sources of UHECRs. We demonstrate that the proposed method drastically improves sensitivity of an experiment by decreasing the minimal required amount of detected UHECRs or the minimal detectable fraction of from-source events several times compared to the approach based on the angular power spectrum. We also test robustness of the neural networks against different models of the large-scale Galactic magnetic fields and variations of the mass composition of UHECRs, and consider situations when there are two nearby sources or the dominating source is not known a~priori. In all cases, the neural networks demonstrate good performance unless the test models strongly deviate from those used for training. The method can be readily applied to the analysis of data of the Telescope Array, the Pierre Auger Observatory and other cosmic ray experiments.