Louis Portales – Phd’s defense – oct. 1, 2020

ector Boson Scattering is among the most sought after types of electroweak process at high-energy collider experiments, as one of the very few processes allowing to probe the quartic gauge boson coupling, and for its close relation to the electroweak symmetry breaking mechanism. These processes are studied through the electroweak production of two bosons associated to two jets (VVjj-EW), that suffer from small cross-sections, an generally from high background contamination from the QCD-driven VVjj production. This thesis presents studies around the WZjj-EW production. Proton-proton collision data collected with the ATLAS experiment at the LHC between 2015 and 2016 at a center-of-mass energy of √s = 13 TeV is used, corresponding to an integrated luminosity of 36 fb^−1. A first part of the work concerns the suppression of pileup jets, not originating from the primary vertex of the process of interest. Pileup jets in the forward regions of the detector cannot be suppressed using tracking information, and the main tagging tool in that region uses full-event topology instead. 

However, that tool can only suppress hadronic pileup jets. With the increasing pileup conditions through Run 2, a non-negligible fraction of jets is reconstructed from stochastic energy deposits in the detector. A new tool combining the original tool to jet shape observables is therefore developed, allowing the identification of these stochastic pile-up jets, and an improved overall tagging efficiency. In a second part, the study of the WZjj production in its fully leptonic decay channels is presented. A multivariate discriminant is developed to optimally separate the WZjj-EW and WZjj-QCD productions. The statistical framework used for the signal extraction is developed around that discriminant, and allowed for the first observation of the WZjj − EW production, with a significance of 5.3 σ, and its cross section measured to be σ(WZjj-EW, fid.) = 0.57 +-0.14(stat.)+-0.07(syst.).