Muon bundle reconstruction with KM3NeT/ORCA using graph convolutional networks

• Published in: Proceedings of 37th International Cosmic Ray Conference — PoS(ICRC2021)
• Main Authors: Reck, Stefan, Eberl, Thomas, Katz, Ulì, Bozza, Cristiano
• Other Authors: Keilhauer, Bianca, Eberl, Thoma
• Format: Conference Object
• Language: English
• Published: SISSA 2022
• Subjects: neural network; muon bundle; KM3NeT; graph network; convolutional network; Orca
• Online Access: https://hdl.handle.net/11386/4877718; https://doi.org/10.22323/1.395.1048; https://pos.sissa.it/395/1048/

KM3NeT/ORCA is a water-Cherenkov neutrino detector, currently under construction in the Mediterranean Sea at a depth of 2450 meters. The project's main goal is the determination of the neutrino mass hierarchy by measuring the energy- and zenith-angle-resolved oscillation probabilities of atmospheric neutrinos traversing the Earth. Additionally, the detector observes a large amount of atmospheric muons, which can be used to study extensive air showers generated by cosmic ray particles. This work describes a deep-learning based reconstruction of atmospheric muons using graph convolutional networks. They are used to reconstruct the zenith angle, the muon multiplicity and the diameter of atmospheric muon bundles. Simulations and measured data from an early four line stage of the detector are used to evaluate the performance. Furthermore, the reconstructions are compared to the ones of classical approaches, and use cases for the indirect study of cosmic ray particles are shown.