Effects of Hadronization, Multiple-Parton-Interactions, Pile-Up and Unfolding Corrections in Multi-jet Events at the ATLAS Detector

In hadron colliders, like the Large Hadron Collider (LHC) located at CERN, near Geneva, Switzerland, jets are produced at a high rate, i.e. many of the observed events contain two, three, four, five, or even six or more jets. Events with such a signature are called multi-jet events and provide fundamental and direct tests of the theory of the strong interaction, the quantum chromodynamics (QCD). Multi-jet events usually represent an important background to searches for other processes and it is therefore important to study them intensively. This is done with the help of the ATLAS detector located at the LHC. In such an analysis, it is essential to understand the so-called multiple-parton-interactions (MPI) and pile-up events in order to be able to make as precise measurements as possible. Multiple-parton-interactions are “spectator-parton” interactions additional to the hard interaction between two partons coming from either of two colliding protons. Pile-up events are proton-proton collisions additional to the hard collision between two protons in one bunch-crossing. Within the ATLAS Multi-jets cross-section analysis, these two effects, as well as effects caused by hadronization, have been studied with the help of the general-purpose Monte Carlo generators PYTHIA 6 and HERWIG. The goal of this thesis is to show the caused effects on several cross section measurements and to compare them qualitatively and quantitatively. Furthermore, one possibility to handle pile-up events, namely by making use of a discriminant called the Jet-Vertex-Fraction (JVF), is described more in detail and evaluated for data measurements. Additionally, detector effects (due to trigger inefficiencies, detector resolutions, etc.) have to be taken into account and one must correct for them. This procedure is called Unfolding and is also described and discussed in detail in the context of the multi-jets cross-section analysis.