Road map for the tuning of hadronic interaction models with accelerator-based and astroparticle data

In high-energy and astroparticle physics, event generators play an essential role, even in the simplest data analyses. As analysis techniques become more sophisticated, e.g. based on deep neural networks, their correct description of the observed event characteristics becomes even more important. Physical processes occurring in hadronic collisions are simulated within a Monte Carlo framework. A major challenge is the modeling of hadron dynamics at low momentum transfer, which includes the initial and final phases of every hadronic collision. QCD-inspired phenomenological models used for these phases cannot guarantee completeness or correctness over the full phase space. These models usually include parameters which must be tuned to suitable experimental data. Until now, event generators have been developed and tuned mainly on the basis of data from high-energy physics experiments at accelerators. The wealth of data available from the latest generation of astroparticle experiments has not yet been fully exploited, and in many cases is not satisfactorily described. Both kinds of data sets are complementary as astroparticle experiments provide sensitivity especially to hadrons produced nearly parallel to the collision axis and cover center-of-mass energies up to several hundred TeV, well beyond those reached at colliders so far. In this report, we provide an overview of state-of-the-art event generators and their tuning, including the most relevant inputs from high-energy accelerator and astroparticle experiments. We present a road map that shows, for the first time, how the unified tuning of event generators with accelerator-based and astroparticle data can be performed.

💡 Research Summary

The paper presents a comprehensive roadmap for the simultaneous tuning of hadronic interaction models using data from both accelerator‑based high‑energy physics (HEP) experiments and modern astroparticle observations. It begins by emphasizing the pivotal role of event generators in simulating the full chain of particle collisions—from the initial state, through hard scattering and parton showers, to hadronisation, decay, and detector response. While accelerator data provide precise measurements of single collisions, they are limited in center‑of‑mass energy (up to a few TeV) and rapidity coverage, leaving the low‑momentum‑transfer, forward‑rapidity region poorly constrained. Astroparticle experiments, on the other hand, probe collisions at center‑of‑mass energies up to several hundred TeV and are especially sensitive to particles produced at very forward rapidities, offering complementary constraints on the same underlying QCD dynamics.

The authors review the current generation of event generators commonly employed in both fields: Pythia 8, EPOS 4 (including the LHC‑R tune), QGSJet II‑1, and Sibyll 2.3d. Pythia 8 relies primarily on perturbative QCD (pQCD) with multiple parton interactions (MPI) and parton‑shower algorithms, supplementing soft physics with phenomenological models. In contrast, EPOS, QGSJet, and Sibyll are built on Gribov‑Regge theory (GRT), modelling hadron–hadron interactions as exchanges of Pomerons and Reggeons, incorporating saturation effects, and extending to nuclear collisions via Glauber‑Angantyr or similar frameworks. The paper highlights how these differing theoretical bases affect each generator’s performance across the accelerator and astroparticle domains.

A dedicated section discusses recent advances in particle transport simulations (e.g., CoREAS, ZHAireS) that are essential for interpreting air‑shower, radio, and Cherenkov signals measured by cosmic‑ray and neutrino observatories. Accurate transport models are required to link the output of hadronic generators to observable quantities such as muon counts, electromagnetic profiles, and radio footprints.

The authors then catalogue the most relevant experimental inputs for a global tuning effort. Accelerator‑based inputs include minimum‑bias and forward‑detector measurements from the LHC (e.g., LHCf, CASTOR), as well as fixed‑target and neutrino‑beam data (NuTeV, MINERvA). Astroparticle inputs comprise extensive air‑shower data from the Pierre Auger Observatory, Telescope Array, KASCADE‑Grande, LHAASO, and muon/neutrino measurements from IceCube/IceTop. For each dataset, the paper outlines systematic uncertainties, the observables that can constrain model parameters, and the mapping between measured quantities and generator outputs.

Current tuning practices are examined, revealing that most automatic tools (Professor, Rivet, Bayesian Analysis Toolkit) are optimized for accelerator data and lack the infrastructure to handle the computationally intensive, multi‑interaction nature of air‑shower simulations. The authors argue for the adoption of Bayesian optimisation, Markov‑Chain Monte Carlo, and surrogate‑model techniques (e.g., neural‑network emulators) to efficiently explore the high‑dimensional parameter space when incorporating astroparticle constraints.

The core contribution is a step‑by‑step roadmap: (1) standardise data formats and metadata for both accelerator and astroparticle measurements; (2) define a set of common observables (e.g., forward particle spectra, muon production depth, neutrino fluxes) that can be directly compared to generator predictions; (3) develop a unified fitting framework that simultaneously evaluates likelihoods from all datasets, employing advanced sampling algorithms to manage computational load; (4) perform sensitivity analyses to identify the most impactful parameters and prioritize them in the tuning; (5) establish a continuous validation pipeline using independent data (e.g., new LHC runs, upcoming IceCube‑Gen2 results) to monitor model performance; and (6) foster an open‑source, community‑driven platform for sharing tunes, tools, and benchmark results.

In conclusion, the paper demonstrates that integrating astroparticle data into the tuning of hadronic interaction models can substantially reduce theoretical uncertainties that currently limit the interpretation of ultra‑high‑energy cosmic‑ray composition, the muon puzzle in extensive air showers, and high‑energy neutrino flux predictions. The proposed roadmap not only bridges the gap between HEP and astroparticle communities but also establishes a sustainable framework for future upgrades as new accelerator facilities (HL‑LHC) and next‑generation astroparticle observatories (CTA, IceCube‑Gen2, GRAND) come online.