Jet Substructure Techniques for the Search of Diboson Resonances at the LHC and Performance Evaluation of the ATLAS Phase-II Inner Tracker Layouts

Jet and track reconstruction are two of the most difficult parts of the event reconstruction of hadron-hadron collisions, and yet they are crucial for the majority of the Large Hadron Collider (LHC) physics analyses. This thesis describes major contributions to jet reconstruction performance and its application in the search for diboson resonances in fully hadronic final states and, in a second part, it presents the design and preparation of the ATLAS tracking system for the future High Luminosity LHC (HL-LHC) physics program. For the reconstruction of boosted hadronic jets, a new object, the Track-CaloCluster, has been developed to unify track and topo-cluster information and maximally benefit, at high transverse momentum, from the energy resolution of the calorimeter and the spatial resolution of the tracker. The combination of the strengths of both the calorimeter and the tracking systems into Track-CaloCluster jets has shown to significantly improve jet substructure performance over a wide range of the kinematic spectrum. This can be directly translated into an increase in sensitivity for the ATLAS physics program, especially in analyses where the jet substructure plays a critical role. Profiting from the improved jet-substructure performance, a novel boosted boson tagging algorithm has been developed and optimised. The optimal working point for the boson tagger is evaluated in order to maximise the sensitivity to the hadronic decay of highly boosted $W$ and $Z$ bosons. The boson identification algorithm has led to improved sensitivity in the search for hadronic decays of boosted vector boson pairs. Compared to topo-cluster jets, Track-CaloCluster jets can provide approximatively 45\% higher significance for 4~TeV $W'$ resonances from the heavy vector triplet model. In the HL-LHC conditions, that will be reached after the LHC upgrade between 2024 and 2026, the usage of the tracking system information will become even more fundamental. To cope with the high luminosity levels, ATLAS foresees important changes and upgrades of the detector which consists of the Phase-II upgrade program. It is crucial that the detector design maintains sensitivity to beyond Standard Model signatures, characterised by high-$p_\mathrm{T}$ leptons, photons, jets, and missing $E_\mathrm{T}$, while improving and extending existing techniques to further expand the experimental physics reach. One of the main components of the ATLAS Phase-II upgrade program is the complete replacement of the Inner Detector with an entirely new tracking system. The new system is required to maintain and even surpass the performance of the current ATLAS Inner Detector, even at the harsh environment after the LHC upgrade. It is imperative to grant high reconstruction efficiency for the different physics objects, provide good vertex definition and pile-up mitigation extending the track-to-vertex matching to the currently-uncovered very-forward region in pseudo-rapidity, improve tracking in dense environment and $b$-tagging performance. The design of the ATLAS Phase-II Inner Tracker and the evaluation of the performance of the candidate layouts are described in detail in the second part of this thesis. The process of defining benchmark performance requirements for the layout candidates to be able to optimise the design of the future detector is discussed in details taking into account many different aspects: tracking performance and their effects on the physics reach, impact on other sub-systems, detector buildability, mechanical properties, etc. The outcome of this process laid the groundwork for the decision of the layout presented in the ATLAS Phase-II Strip and Pixel Technical Design Reports