Search for Supersymmetry in Trilepton Final States with the ATLAS Detector and the Alignment of the ATLAS Silicon Tracker

One of the main goals of the ATLAS detector at the Large Hadron Collider of CERN, a proton-proton collider with a nominal centre-of-mass energy of 14 TeV, is to search for New Physics beyond the Standard Model. A widely favoured Beyond the Standard Model candidate is Supersymmetry (SUSY), which postulates a superpartner with the same quantum numbers, but a spin changed by 1/2 for each Standard Model particle. The first part of this thesis describes a strategy for an early discovery of SUSY using the trilepton signature, with a focus on gravity-mediated SUSY breaking, mSUGRA. The discovery potential for SUSY at the LHC for the case where strongly interacting supersymmetric particles are very massive is critically investigated. A possible choice of triggers for L = 10-31 cm-2 s-1 is suggested by optimising the event yield at intermediate and final selection stages. A novel method to measure the rate of leptons from heavy flavour decays passing isolation requirements by isolating tt events in data is outlined. The task of the ATLAS silicon tracker is to track particles produced in proton-proton collisions in its centre, measuring their momenta and production vertices. The precise knowledge of the silicon tracker module positions and their orientation in space (alignment) down to some microns and fractions of a miliradian in the critical coordinates is of vital importance for large parts of the ambitious ATLAS physics program. In the second part of the thesis, the alignment of the ATLAS silicon tracker using the Robust Alignment algorithm and particle tracks is described. The algorithm is applied to align end-cap A of the pixel detector using cosmic ray particle tracks recorded during its on-surface commissioning in 2006. Finally, about 2M cosmic ray tracks collected by ATLAS in situ in autumn 2008 are utilised to provide a coherent alignment of the entire silicon tracker with the Robust Alignment algorithm.