The so-called LSND anomaly, a 3.8 sigma excess of anti-nu_e events interpreted as originating from anti-nu_mu -> anti-nu_e oscillation, gave rise to many theoretical speculations. The MiniBooNE Collaboration reported inconsistency of this interpretat
ion with the findings from their search for nu_mu -> nu_e oscillations. Yet the origin of the LSND anomaly was never clarified. A critical issue is the prediction of the background anti-nu_e flux that was used in the analysis of the LSND experiment. For this, decisive input comes from pion spectra measured with the HARP large-angle spectrometer under conditions that closely resemble the LSND situation: a proton beam with 800 MeV kinetic energy hitting a water target.
Hadron generation models are indispensable for the simulation and calibration of particle physics detectors. The models used by the Geant4 simulation tool kit are compared with inclusive spectra of secondary protons and pions from the interactions wi
th beryllium nuclei of +8.9 GeV/c protons and pions, and of -8.0 GeV/c pions. We report on significant disagreements between data and model predictions especially in the polar-angle distributions of secondary protons and pions.
Precise measurements of neutrino oscillation parameters and of neutrino-nucleon cross-sections require a good understanding of neutrino beams: flux as a function of energy, transverse beam profile, and flavour composition. For this, hadron production
spectra in proton-nucleus collisions are essential. We report on double-differential inclusive large-angle cross-sections of the production of secondary protons and charged pions, in the interactions with a 5% lambda thick stationary beryllium target, of proton and pion beams with momentum from 3 GeV/c to 15 GeV/c. Our results show cross-sections reported by the HARP Collaboration to be wrong by factors of up to two.
