Study of the non-prompt lepton background in the same charge WW analysis performed in ATLAS at 13 TeV

Vector Boson Scattering (VBS) is a promising process to study the electroweak symmetry breaking mechanism of the Standard Model (SM). A consequence of this symmetry breaking is an additional Boson in the SM: the Higgs Boson. The VBS cross-section is regularized by cancellations of different contributions including the Higgs Boson. Investigations of the energy dependence of the VBS cross-section will show if the SM Higgs Boson fully unitarizes this process as predicted. Compared to other VBS processes, the scattering of two same-sign 𝑊± Bosons (ssWW) has the highest electroweak over strong production cross-section ratio and is therefore well suited to study VBS. Electroweak production of ssWW was first observed last year, with 36.1 $fb^−1$ of the LHC data. The second-largest background and the one with the largest uncertainties to this process is the non-prompt lepton background, which is the focus of this study. Non-prompt leptons are misidentified objects and leptons originating in Photon Conversions or Hadron decays. Due to the characteristics of this background, MC simulations are not accurate. Data-driven techniques are commonly used to estimate the non-prompt background. In this thesis, the 'fake factor method' is used as a starting point to obtain a data-driven assessment of the non-prompt lepton background. Monte Carlo simulations are used to probe the consistency of the method and to study the impact of the event selection and the origin of the non-prompt lepton on the background estimation. A dependence on the event selection and especially on the origin of the non-prompt lepton is observed. Several adjustments to the basic fake factor method are needed in order to obtain reliable predictions. The MC investigations lead to unexpected findings; first, an unusually high amount of non-prompt leptons originating from Photon Conversions is found. Additionally, an issue with the charge compositions of the leptons is noticed. Both issues still need to be understood or confirmed with alternative MC generators and data.