Göttingen 2025 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

AKPIK: Arbeitskreis Physik, moderne Informationstechnologie und Künstliche Intelligenz

AKPIK 3: Machine Learning in Particle- and Astroparticle Physics

AKPIK 3.3: Vortrag

Donnerstag, 3. April 2025, 16:45–17:00, Theo 0.134

Neural Network-Based Event-by-Event Reconstruction of Muon Number from Data of the SD-750 of the Pierre Auger Observatory — •Alina Klingel for the Pierre-Auger collaboration — KIT, Karlsruhe, Deutschland

Ultra-high-energy cosmic rays (∼ 1 EeV) provide a unique opportunity to probe physics beyond the energies of human-made accelerators. At such extreme energies, direct detection is infeasible; instead, these cosmic rays are studied through the particle cascades, or air showers, they generate upon interacting with Earth’s atmosphere. The SD-750 surface detector of the Pierre Auger Observatory records the shower footprint, the spatial distribution of particles and energy deposited on the ground, as time-resolved ground signals. The main advantage of the SD-750 lies in its proximity to the Underground Muon Detector (UMD), allowing for an independent measurement of the muon content of air showers. This setup forms an ideal testbed to develop and benchmark neural network-based estimators for the muon number, even when simulations contain discrepancies. In this contribution, we present a neural network architecture designed to predict the relative muon number in air showers. We aim to shed light on the muon puzzle by cross-calibrating with muon measurements from the UMD.

Keywords: Deep Neural Network; Event-by-Event Reconstruction; Cosmic Rays; Muons; Pierre Auger