Data Analysis in the Direct Dark Matter Search Experiment CRESST and Calculation of the corresponding Limit on the Cross Section of Dark Matter

Abstract

The nature of Dark Matter is one of the most important unsolved questions in astronomy and particle physics. It is proven by different observations and based on different experimental techniques that there is matter in the universe which is not like anything we know from particle physics. The motion of observable matter deviates from the expectation of gravity. Either it could be explained by the introduction of new particles or a modification of the theory of the field of gravitation. This work describes the measurement efforts to be done to detect the hypothetical new particles.

To learn more about this mystery of the universe, dedicated experiments were performed to search for direct or indirect signatures of Dark Matter. A leading one is the CRESST (Cryogenic Rare Event Search with Superconducting Thermometers) experiment. This thesis focusses on the installation and calibration of the muon veto and the following first analysis of a CRESST Dark Matter run taking the muon veto data into account.

In the last part a special method to analyse the data from a low rate experiment is developed. With its help it is possible to compare the results of all direct Dark Matter experiments as objective as possible. For the first time a comparison of different implementations of this analysis method within different background models is done in a 2-dimensional parameter space. Only with this knowledge it is possible to find the best upper limit in presence of an unknown background which is, for example, time and energy dependent.

Um dem Geheimnis der Dunklen Materie auf die Spur zu kommen, gibt es verschiedene Ansätze: Neben den indirekten Methoden der Beschleunigerphysik oder astronomischen Beobachtungen gibt es direkten Suchmethoden. Einen solchen Ansatz verfolgt das Experiment CRESST (Cryogenic Rare Event Search with Superconducting Thermometers). Diese Arbeit beschreibt die Datenanalyse des CRESST Experiments mit dem Schwerpunkt auf des Myonen-Vetodetektorsystem.

Im letzten Teil wird eine Datenanalysemethode vorgestellt, die speziell die Problematik der Statisik seltener Ereignisse berücksichtigt. Um Umweltstörungen noch besser ausschliessen zu können, wurde die Methode hier auf einen 2-dimensionalen Parameterraum erweitert. Damit ist es möglich, den höchsten auszuschliessenden Wert für eine Ereignisrate zu finden, obwohl man eventuell eine unbekannte Quelle für Störereignisse hat, die, zum Beispiel, energie- und zeitabhängig variiert.