Monte-Carlo-Simulation von Top-Paaren in nächstführender Ordnung am ATLAS-Experiment

Due to the probabilistic nature of quantum mechanics, the structure of individual events in particle physics is not calculable ab initio. Therefore, Monte Carlo simulations mark an important tool for understanding experiments in elementary particle physics. Most current Monte Carlo generators employ leading-order matrix elements; these cannot meet the level of accuracy that can be reached by the ATLAS experiment. Although calculations in next-to-leading order have been known for some years, their implementation for generating physically meaningful events is not trivial. The first successful implementation, which defines the standard in this area, is MC@NLO. An alternative approach is the Powheg method, which is examined in more detail in this thesis on the basis of top pair production. The results of both methods are compared and differences are evaluated. It is shown that Powheg represents a viable alternative to MC@NLO that avoids some of its shortcomings. In particular, the method is independent of the parton shower formalism, which allows combining it with the Pythia generator. In this combination, variations of space- and time-like emissions in parton showers are investigated and compared with the results in leading order. The analysis suggests that variations remain substantially valid with matrix elements in next-to-leading order. Nevertheless, significant differences between the two matrix elements are discovered tha t require further study.