NLO event generation for LHC neutrinos and application to flux measurements at FASER
Peter Krack
TL;DR
The paper tackles the challenge of generating accurate, NLO-consistent predictions for LHC forward neutrinos and integrating these into flux measurements at FASER. It implements a LHAPDF-based neutrino flux grid interfaced to the POWHEG-BOX-RES framework, enabling NLO predictions with realistic parton-shower and hadronisation modelling. Through comparisons with GENIE and systematic studies of NLO/QCD, PS, and soft-QCD variations, it quantifies generator-dependent differences and demonstrates a data-driven approach (NNPDF-style) to extract forward neutrino fluxes from FASER measurements using FK-tables. The results show that forward-neutrino predictions are sensitive to forward hadron production modeling and that even a small dataset can discriminate between event generators, with implications for the physics program of the Forward Physics Facility.
Abstract
The LHC generates an intense beam of high-energy neutrinos in the forward direction, whose scientific potential has been left unexploited for many years. The FASER and SND@LHC experiments, operating since 2023, have recently measured LHC neutrinos for the first time. In this contribution we discuss how to produce accurate predictions, including NLO QCD corrections and modern parton shower algorithms, for present and future LHC neutrino experiments, including those at the proposed Forward Physics Facility (FPF). To this end, the energy and rapidity distribution of the LHC neutrinos is encoded in a LHAPDF grid interfaced to the neutrino DIS event generator in the POWHEG-BOX-RES framework. This Monte Carlo tool enables the modelling of differential distributions that are sensitive to hadronic final states, initial- and final-state radiation, and realistic acceptance and selection cuts. As a first application, we deploy this event generator to compute fast-interpolation grids and carry out a first determination of the LHC forward neutrino fluxes directly from FASER data using the NNPDF fitting methodology.
