This thesis describes a data-driven technique for estimating the multijet background to Supersymmetry (SUSY) searches with no leptons using the ATLAS detector at the Large Hadron Collider. The technique is used to estimate multijet distributions in SUSY signal and control regions with 1 fb-1 of sqrt(s) = 7 TeV data collected by ATLAS in 2011. The systematic uncertainty on the estimates is reduced with the development and use of novel event shape triggers. Multijet estimates provided from the technique developed in this thesis are used by the ATLAS collaboration in several different SUSY searches.

This thesis describes searches for new particles predicted by the super symmetry (SUSY) theory, a theory extending beyond the current Standard Model of particle physics, using the ATLAS detector at the CERN Large Hadron Collider. The thesis focuses on searches for stop and sbottom squarks, the SUSY partners of the top and bottom quarks, which are expected to be lighter than the partners of the first and second generation quarks and therefore good candidates for the first evidence of SUSY. It describes novel techniques for estimating and rejecting the Standard-Model backgrounds to searches for these particles. It also includes an independent analysis seeking to constrain the Standard Model ttZ background process, which also represents the first ATLAS search for this rare process at the LHC. The stop squark analysis described, with substantial leading contributions from the author, is the first search for these particles at the LHC to use the jets plus missing transverse energy plus 0-lepton signature and provides the world's best limits on the stop mass for light neutralino LSPs. All in all, the thesis describes three different world-leading analyses in both Standard Model and SUSY physics and therefore represents a major contribution to the field.

The ATLAS experiment at the CERN Large Hadron Collider (LHC) has collected a substantial amount of data to understand the Standard Model of particle physics at higher than previous center of mass energy available and to explore new physics beyond the Standard Model. This dissertation describes observations of charged particle multiplicity distributions in 7 TeV and 900 GeV data as well as searches for new physics with a signature of high energy jets and missing transverse energy using the first few months of data available at the LHC. Multiplicity distributions of charged particle tracks, one of the first observables in high energy collisions were made for a center of mass energy, sqrt (s) = 900 GeV as well as 7 TeV proton-proton collision data. Such distributions help to understand multi-particle production processes. One of the predicted features of multiplicity distribution and its moments is KNO scaling which implies that the shape and moments of the scaled multiplicity distribution is independent of the center-of-mass energy. Although a clear violation of KNO scaling is not observed within the error limits, an indication of such violation is noted. Different models of hadro-production to describe multiplicity distributions are also studied. The Negative Binomial Distribution (NBD) is an often used distribution modeling multiplicity distributions. It has been observed that NBD is satisfied in different types of collisions and over wide range of energies and it was observed that not only the full-phase-space multiplicity distribution can be successfully fitted by the NBD but also the distribution within central pseudo-rapidity intervals. Based on these findings, the model of cluster (or clan) cascading type has been proposed. Although, a good NBD fit can be obtained, it is observed for hadronic interactions that the presence of two weighted NBD or Double NBD (DNBD) components, one corresponding to soft production and the other to semi-hard one (mini-jets) seems to fit the data better for broad pseudo-rapidity ranges. It was found that the soft component follows KNO scaling while the semi-hard component does not. The proton-proton collision data at LHC has been analyzed to test the NBD and DNBD parametrization and test the energy dependence of the fitted parameters. It is well understood that the Standard model of Particle physics in incomplete. Our knowledge of cosmology also leaves several crucial questions unanswered, one of them being the composition of the dark matter that has been indirectly observed through astronomical observations. The primary objectives for constructing the Large Hadron Collider has been to solve these problems through the discovery of the Higgs particle and to find new physics processes that predict the production of massive non-interacting stable particles. Searches of such new physics producing heavy stable particles in its final state has been performed. Finding such signals would provide direct observation of a dark matter candidate particle. The exclusive event topology of two high energy jets and missing transverse energy has been explored to perform the above search, using ATLAS detector data. A summary of results of these preliminary searches in comparison with theoretical predictions has been presented. In order to understand and discriminate any new physics, a clear and coherent understanding of the detector response is crucial. Moreover, a detailed knowledge of the behavior of known standard model phenomena is required. A detailed description of the ATLAS detector and several important calibration techniques is discussed and their results summarized. Estimates of Standard Model physics, contributing to irreducible backgrounds to dark matter searches is presented in detail. One such physics process constituting an irreducible background is the production of the Z boson decaying into two neutrinos ([nu]) with associated jets. The observation of these events directly from data is an impossible task due to the non-interacting nature of the neutrinos. An estimate of the production cross section of these process is estimated using observations of photon ([gamma]) plus jet events and theoretical predictions. Estimated numbers for the [gamma] plus 2,3,4 and more associated jets production with uncertainties has been summarized.

The Standard Model of particle physics, despite being extremely successful, is not the ultimate description of physics. The nature of dark matter is not well described, unification of the forces is not achieved and the theory is plagued by a hierarchy problem. One of the proposed solutions to these issues is supersymmetry. This thesis describes numerous searches for supersymmetry carried out using the ATLAS detector at the Large Hadron Collider. In scenarios where R-parity is conserved, supersymmetric final states contain large amounts of missing transverse energy. Furthermore, should supersymmetry correctly describe Nature, the scalar partners of the third generation quarks might be the lightest scalar quarks. The searches reported here exploit these possibilities and make use of signatures which are rich in missing transverse energy and jets coming from heavy flavour quarks. Searches are carried out for direct pair production of third generation scalar quarks as well as gluino-mediated production of these particles. A data driven technique to estimate the backgrounds coming from multijet production is described and shown to work in analyses targeting heavy flavour quarks. No significant excesses are observed in a number of analyses. In each case limits are set on the allowed masses of supersymmetric particles in a variety of phenomenological models and in specific supersymmetry breaking scenarios.

This thesis represents one of the most comprehensive and in-depth studies of the use of Lorentz-boosted hadronic final state systems in the search for signals of Supersymmetry conducted to date at the Large Hadron Collider. A thorough assessment is performed of the observables that provide enhanced sensitivity to new physics signals otherwise hidden under an enormous background of top quark pairs produced by Standard Model processes. This is complemented by an ingenious analysis optimization procedure that allowed for extending the reach of this analysis by hundreds of GeV in mass of these hypothetical new particles. Lastly, the combination of both deep, thoughtful physics analysis with the development of high-speed electronics for identifying and selecting these same objects is not only unique, but also revolutionary. The Global Feature Extraction system that the author played a critical role in bringing to fruition represents the first dedicated hardware device for selecting these Lorentz-boosted hadronic systems in real-time using state-of-the-art processing chips and embedded systems.