Enhancing Events in Neutrino Telescopes through Deep Learning-Driven Super-Resolution

TL;DR

This paper introduces a deep learning super-resolution method that predicts additional photon hits in neutrino detectors, enhancing the angular resolution of muon events and potentially improving overall neutrino event reconstruction.

Contribution

It proposes a novel deep learning approach to simulate virtual optical modules, boosting reconstruction accuracy in neutrino telescopes beyond traditional methods.

Findings

Improved angular resolution for muon events in ice-based detectors.

Method extends to water-based neutrino telescopes and various event types.

Enhances reconstruction accuracy using virtual optical modules.

Abstract

Recent discoveries by neutrino telescopes, such as the IceCube Neutrino Observatory, relied extensively on machine learning (ML) tools to infer physical quantities from the raw photon hits detected. Neutrino telescope reconstruction algorithms are limited by the sparse sampling of photons by the optical modules due to the relatively large spacing ($10-100\,{\rm m})$ between them. In this letter, we propose a novel technique that learns photon transport through the detector medium through the use of deep learning-driven super-resolution of data events. These ``improved'' events can then be reconstructed using traditional or ML techniques, resulting in improved resolution. Our strategy arranges additional ``virtual'' optical modules within an existing detector geometry and trains a convolutional neural network to predict the hits on these virtual optical modules. We show that this technique improves the angular reconstruction of muons in a generic ice-based neutrino telescope. Our results readily extend to water-based neutrino telescopes and other event morphologies.