Position reconstruction and data quality in XENON

Abstract

The XENON collaboration operates a dark matter detector (XENON100) at Gran Sasso, Italy. A new detector (XENON1T) is under construction. This thesis reports on the working and implementation of a position reconstruction method developed for XENON1T, the definition of several data quality cuts for XENON100 and the first physics analysis performed with a new data processor for XENON1T, called PAX. The detectors allow the position of an event to be reconstructed in three dimensions. The χ 2 γ position reconstruction algorithm is physically motivated and provides a per- event error estimation. It is implemented in the data processor for XENON1T and tested using simulations in a XENON100 configuration. The new implementation performs better than its main competitor, a Neural Network. Two data quality cuts are defined to reject poorly position reconstructed events for the latest calibration run of XENON100. Both cuts are redefined compared to their previous versions to cope with a higher noise level in the XENON100 detector. A third cut on the ratio of light seen in the top and bottom of the detector (asymmetry) is also redefined. The new data processor PAX is used to study and classify a group of previously unseen events with very low asymmetry. The events have a very unusual combination of properties implying that they may be caused by scintillation light in the liquid xenon, a process very unlikely to happen in XENON100. Further research will be needed to test this hypothesis.