Measurement of cosmic-ray Helium flux in extended acceptance with the CALET experiment

State-of-the-art detectors operating in space have paved the way for high-precision direct measurements of galactic cosmic-ray (CR) spectra. Recent results for proton and helium nuclei have shown unexpected spectral features that deviate from the single power law predicted by the standard CR model. From a few hundred GeV to a few TeV, a hardening of the spectral index, i.e. an enhancement of the flux, has been measured by several experiments (e.g. AMS-02, PAMELA, DAMPE, and CALET) using different experimental techniques (magnetic spectrometers, calorimeters). At tens of TeV, both the DAMPE and CALET calorimeters have recently observed a flux softening. These results have pushed the boundaries of direct measurements of CR spectra, helping to shed light on the acceleration and propagation mechanisms of CRs in the Galaxy. Nevertheless, the high energy region of the spectra, from tens to hundreds of TeV, still suffers from significant uncertainties, mainly due to the very limited statistics. In this context, the main objective of this thesis is to improve the statistical precision of the CR helium flux measurement with CALET data, focusing on the high-energy region. This is achieved by extending the fiducial geometrical acceptance of the present analysis. The CALorimetric Electron Telescope (CALET) is a multi-purpose space-based experiment that has been acquiring data onboard the International Space Station (ISS) since mid-October 2015. The mission is sponsored by JAXA (Japan Aerospace Exploration Agency) with the collaboration of ASI (Italian Space Agency) and NASA (National Aeronautics and Space Administration). The CALET instrument consists of three sub-systems. The Total AbSorption Calorimeter (TASC), a deep homogeneous calorimeter with an equivalent thickness of 27 radiation lengths (X0 ), measures the particle energy. The IMaging Calorimeter (IMC) is a sampling calorimeter with an equivalent thickness of 3 X0 , primarily designed to visualize the particle trajectory and its early shower profile. The CHarge Detector (CHD) is a two-layer hodoscope for identification of nuclear species over a wide dynamic range, up to Z = 40. The first chapters of this thesis provide both an overview of the physics of cosmic-rays, focusing on the galactic component investigated by the CALET telescope, and a detailed description of the instrument. The event reconstruction procedure for in-flight and simulated data is described as well. The core chapters of the thesis are dedicated to the newly developed analysis strategy for helium flux measurements with higher statistical accuracy. The main novelty is the implementation of a multivariate analysis based on Boosted Decision Trees (BDT) to improve the analysis performance, while mitigating the issues arising from the extension of the geometric acceptance. The stability of the unfolding procedure for inferring the primary energy from the fraction of energy deposited in the calorimeter, is also validated using both simulated and in-flight data. Finally, the helium flux measurements in fiducial and enlarged acceptances are presented showing a statistical gain up to 60%, and the consistency with the previously published analysis.