Measuring the charged pion and kaon contributions to the T2K neutrino beam in neutrino and antineutrino running modes using ν μ and ν ̄ μ CC inclusive selections in the ND280 detector

Abstract

Tokai to Kamioka (T2K) is a long-baseline experiment designed to study neutrino oscillations using an (anti)neutrino beam. The beam is created by colliding 30 GeV protons with a stationary graphite target. Secondary hadrons, primarily pions and kaons, are selected using magnetic horns and they subsequently decay to final states that include neutrinos. Depending on the magnetic horn current, a neutrino or antineutrino enhanced beam can be produced. It is important to understand the flux and flavour content of the beam, in order to maximise sensitivity of the experiment. The T2K experiment uses flux models constrained by external measurement from the NA61/SHINE experiment [1] [2]. Despite this, an in-situ measurement is required to confirm the flux model derived from the NA61/SHINE data, account for time variations and any other differences between the modelled and actual beamline. This thesis used inclusive charged-current (CC) νµ and ¯νµ selections to obtain (anti)muon samples binned in reconstructed momentum. A χ 2 fit, which took into account all statistical and shape-only systematic errors, was performed on the (anti)muon momentum distributions to determine the parent hadron yields. The fit was performed on each neutrino and antineutrino T2K data run separately, and to the entire neutrino data set. The data to Monte Carlo ratio for the parent hadron yields, with shape-only uncertainties applied, was found to be 1.080 ± 0.039 for π + and 0.981 ± 0.080 for K+ in the neutrino enhanced data set, and 1.113±0.124 for π +, 1.089±0.182 for K+, 0.980±0.069 for π − and 0.880 ± 0.230 for K−, in the antineutrino enhanced run 6 data set. This shapeonly analysis provides a comparison of neutrino parent yields relative to each other. This does not constitute an absolute measurement of the yields as that would limited by the normalisation uncertainties on the cross-section models, at around 10%. The normalisation uncertainties have been added after the fit and the resulting distribution is consistent with the nominal distribution at the 1σ level. No significant time variation was found, and results were in agreement with the NA61 model at around the 2σ level for neutrino and antineutrino beams.