Measurement of Higgs boson fiducial cross sections with the CMS detector and electromagnetic reconstruction with the high-granularity endcap calorimeter

This PhD thesis presents a comprehensive characterisation of the Higgs boson in the H → ZZ → 4ℓ decay channel, commonly known as the four-lepton channel, using proton-proton collisions at a centre-of-mass energy of 13 TeV recorded with the CMS experiment during the Run 2 of the CERN LHC. The four-lepton channel is regarded as the golden channel of Higgs physics due to its clear peak over an almost flat background, the large signal-to-background ratio, a fully reconstructible final state, and the synergy with the highly performing lepton reconstruction of the CMS detector. Cross section measurements are one of the best methods to study the Higgs boson properties, probing its couplings to other particles and testing theoretical predictions. In this thesis, cross sections are obtained by removing detector effects from data and measured in a fiducial phase space defined to match the experimental acceptance closely. This methodology ensures maximal model independence and reinterpretability for the results. Cross sections are measured both inclusively and in bins of 32 single- and double-differential observables, providing insights into the production and decay of the Higgs boson, the tensor structure of the vertex between the Higgs boson and the Z bosons, as well as QCD effects. The constraint from data of the non-resonant ZZ→4ℓ background is also included. The set of results is completed by the constraints of the trilinear self-coupling of the Higgs boson and its couplings to bottom and charm quarks. All results are consistent with the theoretical predictions of the standard model of particle physics. Looking ahead, with the Run 2 concluded and Run 3 underway, this thesis introduces a novel method for estimating the reducible background of the four-lepton channel that will be used in forthcoming analyses. This method can also be extended to all channels with leptonic final states. The reducible background comprises non-prompt leptons and other particles misidentified as leptons, which can mimic the signal signature. The proposed strategy explores the possibility of modelling this source using a novel approach and reducing the sizeable systematic uncertainty typical of the current methods, which will be a limiting factor during the next phase of the LHC. The High-Luminosity LHC (HL-LHC) aims to increase the integrated luminosity by a factor of 10 beyond the LHC's design value, opening new horizons for discoveries and precision physics. In order to cope with the large number of simultaneous collisions per bunch crossing, known as pileup, and sustain the high radiation dose, the CMS experiment foresees the complete replacement of the endcap calorimeter. The new High-Granularity endcap CALorimeter (HGCAL) will be a silicon-based sampling calorimeter, offering the possibility of performing calorimetry with tracker-like granularity. This thesis contributes to the development and reassessment of the electron and photon offline reconstruction for the HGCAL. The first contribution regards cleaning electromagnetic showers from spurious contaminations resulting from the high-pileup environment, which degrades the properties of reconstructed electromagnetic objects. The second contribution focuses on electron reconstruction. An electron can start showering while traversing the inner tracker before reaching the calorimeter. This effect, combined with the 3.8 Tesla of the CMS magnet, leads the electron energy to be spread in several clusters around the primary one. These contributions should be clustered together to reconstruct the original electron. A purely geometrical algorithm currently performs this procedure, and its performance is assessed in the HGCAL for the first time. Additionally, this thesis proposes a new, dedicated algorithm based on Deep Neural Networks explicitly tailored for the new calorimeter.