Experimental proof of principle of the Neutrino Tagging at NA62

This thesis work objectives belong to the framework of neutrino physics at accelerators. In this context, an innovative technique is proposed, that will substantially reduce the systematic uncertainties linked to the neutrino oscillation studies: the neutrino tagging technique. The neutrino tagging technique proposes to instrument a neutrino beam line with silicon trackers to kinematically reconstruct properties of individual beam neutrinos produced in $\pi^\mp \to \mu^\mp \overline{\nu_{\mu}}, K^\mp \to \mu^\mp \overline{\nu_{\mu}}$ decays. As a result, the initial neutrino flux is precisely determined and the individual neutrino energy can be reconstructed with a resolution better than 1%. Moreover, based on time and angular coincidence, the neutrinos kinematically reconstructed by the trackers can be individually associated to the neutrinos interacting in the neutrino detector, such that the precise measurement of their properties can be used for physics analyses (e.g. oscillations, cross-section). The main aim of this work is to demonstrate the feasibility of the neutrino tagging technique. The proof of principle of the neutrino tagging method has been performed using the NA62 experiment at CERN as a miniature neutrino experiment: its intense kaon beam copiously produces neutrinos when decaying as $K^+ \to \mu^+ \nu_{\mu}$, its spectrometers acts as tagger for the charged particles, and its electromagnetic calorimeter serves as neutrino detector. The implementation of the neutrino tagging at a Long Baseline Neutrino Experiment is described in detail, as well as the blind analysis performed on NA62 data. Two tagged neutrino candidates are found in the signal region, proving the feasibility of this technique. Finally, the data analysis on the data from a test beam performed on modules based on the NA62 GigaTracKer technology, aimed to study the time resolution budget in silicon pixel detectors, is presented.