Constraining neutrino flux predictions with hadron production data: the NA61/SHINE measurements for the T2K experiment.

From hypothetical particles in the 30s, neutrinos have turned into the first hint for new physics beyond the Standard Model in the late 90s. The observation of a new phenomenon, referred to as neutrino oscillations, in which a neutrino of a given flavor transforms into a neutrino of another flavor after traveling over a sufficiently long distance, actually revealed properties of the leptonic sector of the Standard Model that were not described so far: neutrinos have mass, leptons mix, and consequently, there might be CP violation in the leptonic sector. The T2K experiment at Tokai, Japan, will allow us to probe neutrino oscillations with an unprecedented precision on the oscillation parameters in the three-neutrino model. The experiment is designed to perform a precise measurement of the numb to nue oscillation driven by the atmospheric mass squared splitting |∆ m^2_{23}| and the so far unknown mixing angle theta_{13}. To achieve a sensitivity down to sin^2(2 theta_{13})=0.006 at 90% CL and a precision of delta(∆m^2_{23})=10^{-4} eV^2, the T2K experiment needs a high intensity neutrino beam and a precise prediction of neutrino fluxes. The first point is achieved by using a high intensity proton beam on target with a design beam power of 0.75 MW, the second point by tuning the original predictions of the T2K beam simulation with hadron production data measured by the NA61/SHINE collaboration at the CERN SPS. The latter performed measurements of p+C interactions at 31 GeV/c beam momentum on both a thin (4% of the interaction length) target and a full-size (1.9 interaction length) replica of the T2K production target. The current neutrino flux predictions in T2K are based on the usual approach, which consists in tuning the particle multiplicities of the simulated primary interactions with inclusive hadron production cross section measurements. The goal of this thesis is to develop a complementary approach in which the tuning of the flux predictions is done at the level of both primary and secondary interactions by using long target measurements. Such a tuning would actually constrain a larger fraction of the neutrino flux, with a potentially better control of systematic uncertainties. The first chapter of this thesis is an overall introduction to neutrino physics and to the phenomenology of neutrino oscillations. The original beam simulation of the T2K experiment and corresponding flux predictions from different sources are presented in Chapter 2, together with modifications implemented in order to express the flux predictions in terms of hadron production measurements off thin and long targets. These studies stressed the importance of the long target measurements to constrain neutrino flux predictions. The first part of Chapter 3 is dedicated to the estimation of the statistics of the NA61/SHINE long target measurements based on the T2K physics requirements. The second part of the chapter presents the analysis of the NA61/SHINE 2007 long target data, which provides a direct comparison of the total pi+ production simulated by the T2K hadron production model to the measured production over the full-size replica of the T2K target. In Chapter 4, a first method is proposed to use the NA61/SHINE long target data within the T2K beam simulation and re-weight the neutrino flux predictions accordingly. Finally, the recent results of the T2K nue appearance analysis are reported in Chapter 5, in particular, the usage of the NA61/SHINE thin target data for the neutrino flux predictions.