Estimation of cosmic ray mass by correlating muon signals extracted from surface detector stations of the Pierre Auger Observatory using neural networks

To fully understand the acceleration and propagation physics of cosmic rays at the highest energies, it is necessary to have an accurate knowledge of their mass composition. However, since information on the mass and energy is degenerate and a direct measurement of the mass is impossible, obtaining information about the composition is non-trivial. One of the observables that break this degeneracy is the number of muons produced during the particle cascade induced by a cosmic ray interacting with the Earth's atmosphere.

The Pierre Auger Observatory is the largest cosmic ray detector on Earth. One of its central components is its surface detector array consisting of uniformly spaced hybrid detector stations. These stations measure time traces of particles produced in the shower cascade.

This contribution presents a method of estimating the number of muons from these signals using machine learning techniques. The signal trace from a single station and a fixed set of shower parameters are used as input of a neural network to infer the fraction of the signal that is due to muons. The fractions obtained from several stations can be spatially correlated to estimate the mass of the primary cosmic ray. ... mehrEventually, the estimated muon content of recorded air showers will be presented including a study of systematic uncertainties.