Analysis and estimation of uncertainties in H->ττ and Z->ττ analyses with the CMS experiment

Abstract

The LHC Run II will provide with 13 TeV almost double the center of mass energy than the previous run. A major focus at the LHC lies on the measurement of the properties of the so-called "Higgs boson". After discovering it in 2012 the LHC Run II now offers the possibility to solidify this model for further research. Since the Higgs boson cannot be detected directly, it is necessary to analyse the end products of its decay channels. An essential tool is the simulation of proton proton collisions to generate events, which offer a good basis to test mathematical models before being applied to the real life data delivered by the CMS detector at LHC. This simulation is based on Parton Distribution Functions (PDFs), describing the impulse share of each parton (quarks or gluons) in a proton. This thesis is dedicated to the question, which effects small variations in the PDFs have on the results of the simulation of measurable observables. Especially the H-> tau tau and Z->tau tau decay channels are subject to scrutiny. This was done by determining the PDF Uncertainties of simulated data (using MADGRAPH and POWHEG event generators) and exploring their effects on observables such as the visible mass mvis, as recommended by the PDF4LHC group. This thesis concludes that the resulting uncertainties are generally in the range of 0.18% and 1.3%. Concerning the estimated uncertainties of each sample, PDF uncertainties might play a slightly more significant role in Drell-Yan processes than in Vector Boson fusion or gluon fusion. In contrast stand the uncertainties deriving from scale variation. These turn out significantly higher, especially for gluon gluon fusion, were in the most extreme case values of up to 12% are estimated. This suggests, that Scale uncertainties have a larger impact on events than PDF uncertainties. Ratio plots visually show that the PDF uncertainties are constantly smaller than the statistical error of the distribution itself.