Uncertainties in astrophysical gamma-ray and neutrino fluxes from proton-proton cross-sections in the GeV to PeV range

TL;DR

This paper investigates how uncertainties in proton-proton cross-section models affect predictions of gamma-ray and neutrino fluxes from cosmic ray interactions, impacting the identification of astrophysical sources.

Contribution

It introduces the implementation of various hadronic interaction models into the CRPropa code and assesses their impact on flux predictions in astrophysical environments.

Findings

Flux predictions vary by a factor of about 2 depending on the HI model used.

Different HI models trained on observational data produce significantly different gamma-ray and neutrino fluxes.

The choice of hadronic interaction model is crucial for accurate multimessenger astrophysics analyses.

Abstract

The identification of Cosmic Ray (CR) sources represents one of the biggest and long-standing questions in astrophysics. Direct measurements of cosmic rays cannot provide directional information due to their deflection in (extra)galactic magnetic fields. Cosmic-ray interactions at the sources lead to the production of high-energy gamma rays and neutrinos, which, combined in the multimessenger picture, are the key to identifying the origins of CRs and estimating transport properties. While gamma-ray observations alone raise the question of whether their origin is hadronic or leptonic, the observation of high-energy neutrino emission directly points to the presence of CR hadrons. To identify the emission signatures from acceleration and transport effects a proper modeling of those interactions in a transport framework is needed. Significant work has been done to tune the production cross sections to accelerator data and different models exist that put the exact evolution of the Monte-Carlo generated showers into a statistical approach of a probabilistic description of the production of the final states of the showers relevant for astrophysical observations. In this work, we present the implementation of different hadronic interaction (HI) models into the publicly available transport code CRPropa. We apply different descriptions of the HI, trained on observational data in different energy regimes to a nearby, giant molecular cloud and the Galactic diffuse emission. In this case, the resulting gamma-ray and neutrinos fluxes can differ by a factor $\sim 2$ dependent on the choice of the HI model.