Speaker
Description
Ultra-high-energy photons are expected as by-products of interactions between ultra-high-energy cosmic rays (UHECRs) and the cosmic microwave background, as well as from UHECR interactions with galactic matter and other, more exotic processes. Despite these various production mechanisms, the prospect is that the diffuse photon flux is low enough to prevent direct detection. Consequently, photon searches must rely on large ground-based detector arrays, suitable for this energy regime. In this contribution, we present a method for photon-hadron discrimination that applies deep learning algorithms to detector simulations within the context of the Pierre Auger Observatory. Our method correlates information from both the Surface Detector (SD) and the Underground Muon Detector (UMD). The SD, composed of water-Cherenkov detectors, measures the secondary particles arriving at the ground from the muonic and electromagnetic components of extensive air showers. In contrast, the UMD counters, shielded by soil, measure muons above $\sim 1$ GeV. We chose graph neural networks (GNNs) for their effectiveness in handling the discrimination task, allowing for an easy and flexible correlation of information from the SD and UMD. This approach is particularly suitable for handling the irregular structures found in SD and UMD configurations, where stations and counters may be missing due to technical issues, as GNNs are designed to manage such graph-based inputs. Using simulations, the performance estimates indicate that the method has strong potential for identifying photons.
