Investigation to Separate Cosmic-Ray Air Showers by Primary Particle Mass using Knowledge of Multiple Air-Shower Observables

Cosmic rays, high-energy astrophysical particles, interact with nuclei in Earth’s atmosphere to produce extensive air showers. Air-shower observables, such as the depth of shower maximum and the total muon content of the shower, are statistically related to the primary cosmic-ray mass. Yet each observable is intrinsically limited in its ability to separate primary cosmic rays by mass on a per-event basis. These limitations can be softened by studying multiple mass sensitive observables in parallel. Air-showers simulated at the locations of both the IceCube Neutrino Observatory at the South Pole and the Pierre Auger Observatory in Argentina were used to obtain knowledge of multiple air-shower observables. For each location, both a Fisher linear discriminant analysis and a gradient boosted decision tree were used to study the separation between populations of different primary particles. We find combining knowledge of all studied observables provides the best avenue for mass separation, where the combination of muon and depth of shower maximum observables contribute most to this separation. Hence, our results motivate the precise reconstruction of these observables for future air-shower observatories.