Simulation chain and signal classification for acoustic neutrino detection in seawater

TL;DR

This paper presents a comprehensive Monte Carlo simulation chain for acoustic neutrino detection in seawater, including signal classification techniques that significantly improve background suppression by leveraging the unique propagation pattern of neutrino signals.

Contribution

It introduces a detailed simulation framework from neutrino interaction to signal classification, incorporating real background data and refraction effects, enhancing detection capabilities.

Findings

Background suppression improved by nearly two orders of magnitude.

Geometrical shape of sound propagation is a key feature for classification.

Refraction effects influence the neutrino signal signature.

Abstract

Acoustic neutrino detection is a promising approach to extend the energy range of neutrino telescopes to energies beyond $10^{18}$\,eV. Currently operational and planned water-Cherenkov neutrino telescopes, most notably KM3NeT, include acoustic sensors in addition to the optical ones. These acoustic sensors could be used as instruments for acoustic detection, while their main purpose is the position calibration of the detection units. In this article, a Monte Carlo simulation chain for acoustic detectors will be presented, covering the initial interaction of the neutrino up to the signal classification of recorded events. The ambient and transient background in the simulation was implemented according to data recorded by the acoustic set-up AMADEUS inside the ANTARES detector. The effects of refraction on the neutrino signature in the detector are studied, and a classification of the recorded events is implemented. As bipolar waveforms similar to those of the expected neutrino signals are also emitted from other sound sources, additional features like the geometrical shape of the propagation have to be considered for the signal classification. This leads to a large improvement of the background suppression by almost two orders of magnitude, since a flat cylindrical "pancake" propagation pattern is a distinctive feature of neutrino signals. An overview of the simulation chain and the signal classification will be presented and preliminary studies of the performance of the classification will be discussed.