Optimising triggers and jet reconstruction for Higgs measurement with the ATLAS experiment

Since observation of the Higgs boson, one of the fundamental priorities of the ATLAS experiment is to precisely measure properties of this particle, allowing evaluation of the Standard Model (SM) of particle physics. Associated Higgs production with a top-quark pair (ttH) is particularly useful as it can directly determine the absolute value of the top-Yukawa coupling. Correctly identifying events containing interactions with the Higgs is challenging: among all ttH final states, that in which a Higgs decays to a bb-pair and and the top quarks each decay hadronically to jets has the largest branching ratio, but also the least signal purity. The overwhelming majority of backgrounds within ATLAS come from non-resonant production of multijet events. The other significant background in this search comes from tt+ jets events. Distinguishing signal events from background events based around the identification and classification of jets becomes increasingly difficult as pile-up increases; proposed increases in luminosity at the LHC will make this worse. Optimising the ATLAS trigger and jet reconstruction are required to combat this effect. In the trigger, I investigate the viability of tracking with jets, opening new avenues to improving identification. Developments in jet reconstruction, through access to jet substructure (JSS), can improve flavour-tagging. Limitations arise due to contributions from Non-Global Logarithms (NGL), a problem circumnavigated by utilising soft drop grooming. By analysing JSS variables in flavour-tagged soft drop-groomed jets I am able to identify optimal grooming strengths and variables for discriminating between different jet flavours. In particular, I find that applying soft-drop grooming to tau-flavoured jets improves correlation between the jet mass and the mass of tau leptons. Comparing reconstructed jets with truth jets at different grooming strengths also allows me to evaluate the validity of reconstructed jets in representing real physics. Finally, I use all this to compare results from Monte Carlo (MC) simulation of tt, ttH and dijet events, as well as real data from the ATLAS detector in order to optimise discrimination between Higgs events and background. These processes display potential for optimising Higgs measurements with flavour-tagging improvements through JSS analysis in soft drop-groomed small-jets.