Calorimeter-Based Triggers at the ATLAS Detector for Searches for Supersymmetry in Zero-Lepton Final States

Abstract

This thesis consists of three closely related parts. An analysis of data recorded by the ATLAS detector in 2010 in proton-proton collisions at a center-of-mass energy of 7 TeV with an integrated luminosity of 33.4 pb-1 is performed, searching for supersymmetric final states containing jets and missing transverse energy and no electrons or muons (zero-lepton channel). No excess over the Standard Model background expectation is observed. Using the CLs and PLR methods, exclusion limits are set in a minimal supergravity model with tan β = 3, A0 = 0 GeV and μ > 0 and in a simplified supergravity model. These considerably extend the excluded parameter ranges from earlier experiments. The analysis validates the official analysis carried out within the ATLAS Supersymmetry group and additionally takes into account the uncertainty from pile-up effects.The rates and efficiencies of triggers based on combined signatures with jets plus missing transverse energy in ATLAS are studied, which are the primary triggers for the search for Supersymmetry in the zero-lepton channel. Methods to measure the efficiencies on data are tested and optimized to obtain sufficient statistics. For the measurement of the efficiencies in data collected in 2010 and 2011, the bootstrap method is applied to correct for the sample trigger bias. Different sample triggers based on jets and missing transverse energy are compared and their efficiencies are measured. A reweighting approach is studied and used to correct for the bias from the propagation of the uncertainties in the bootstrap method. The resulting efficiency estimates for the primary triggers allow to determine the onset of the plateau of the two-dimensional turn-on curves and are input to the official analyses in the ATLAS Supersymmetry group in 2010 and 2011.A universal model is developed to describe the contribution of fake missing transverse energy from resolution effects to the rates of missing transverse energy triggers as function of the level of in-time pile-up, i.e. the number of concurrent proton-proton interactions. The input parameters to tune the model to the properties of the ATLAS triggers are determined, and the model predictions are compared to measurements of trigger rates in ATLAS. Good agreement is found for missing transverse energy triggers with low thresholds for which the rates are dominated by resolution effects, whereas the rates for higher thresholds are underestimated due to additional sources of fake and real missing transverse energy which are not incorporated in the model.