Testing cosmic ray composition models with very large volume neutrino telescopes

TL;DR

This paper investigates how different models of cosmic ray composition influence the expected neutrino detection rates in large volume neutrino telescopes, highlighting current discriminative capabilities and future improvements.

Contribution

It analyzes the impact of primary cosmic ray composition models on neutrino telescope observations and assesses the potential for current and future detectors to distinguish between these models.

Findings

Current IceCube data can partially discriminate between composition models.

Next-generation telescopes will improve sensitivity to model differences.

Uncertainties in cosmic ray physics significantly affect neutrino flux predictions.

Abstract

The composition in terms of nuclear species of the primary cosmic ray flux is largely uncertain in the knee region and above, where only indirect measurements are available. The predicted fluxes of high-energy leptons from cosmic ray air showers are influenced by this uncertainty. Different models have been proposed. Similarly, these uncertainties affect the measurement of lepton fluxes in very large volume neutrino telescopes. Uncertainties in the cosmic ray interaction processes, mainly deriving from the limited amount of experimental data covering the particle physics at play, could also produce similar differences in the observable lepton fluxes and are affected as well by large uncertainties. In this paper we analyse how considering different models for the primary cosmic ray composition affects the expected rates in the current generation of very large volume neutrino telescopes (ANTARES and IceCube). We observe that, a certain degree of discrimination between composition fits can be already achieved with the current IceCube data sample, even though in a model-dependent way. The expected improvements in the energy reconstruction achievable with the next generation neutrino telescopes is be expected to make these instruments more sensitive to the differences between models.