The CRESST experiment (Cryogenic Rare Event Search with Superconducting Thermometers) is a cutting-edge direct detection dark matter experiment utilizing cryogenic detectors. CRESST aims to detect dark matter particles, specifically Weakly Interacting Massive Particles (WIMPs), by using scintillating crystals, made in the default configuartion from calcium tungstate (CaWO4) and operated at millikelvin temperatures. This experimental setup allows for the precise measurement of energy deposited by potential dark matter interactions.

 

CRESST detectors are equipped with superconducting transition-edge sensors (TES), which measure both phonon (heat) and scintillation light signals produced by particle interactions within the crystals. This dual-signal approach helps in distinguishing between different types of interactions, thereby enhancing the experiment's sensitivity to potential dark matter signals.

 

Each CRESST detector module consists of a ~25g CaWO₄ crystal coupled to a light detector. When a particle interacts with the crystal, it produces a small amount of heat and scintillation light. The heat is measured by the TES, while the light is detected by a separate light detector. The ratio of the phonon to light signal helps differentiate between nuclear recoils (possible dark matter interactions) and electron recoils (background events).

 

The CRESST experiment is located at the Gran Sasso National Laboratory (LNGS) in Italy, benefiting from the lab's deep underground environment which significantly reduces cosmic ray background noise. This low-background setting, combined with the high sensitivity of cryogenic detectors, enables CRESST to explore new parameter space in the search for dark matter.

 

More information about CRESST can be found here.

For analyses very close to a low energy threshold, a detailed understanding of the raw data and the causes of, and influences on, the noise baseline is of utmost relevance. My group is therefore particularly concerned with the analysis of raw data. One focus here is the development of algorithms based on machine learning. In preparation for a large-scale expansion of CRESST, existing AI algorithms will also be further developed for automated data processing.