Developing the statistical tools for dark matter discovery

The dark matter group at UCL plays a key role in the world-leading LZ experiment while the group at Stockholm does so within the directly competing XENONnT experiment. Both seek the common goal of exploring orders of magnitude of new parameter space for the well-motivated dark matter WIMP candidate (weakly interacting massive particle) as well as providing a number of other physics searches of international priority. 

Flamedisx (fast likelihood analysis in more dimensions using liquid xenon TPCs) is a novel statistical inference tool developed initially within the Stockholm group. It enables likelihood (data probability given a certain physics model) evaluation via a convolution of analytic probability elements associated with different parts of a detector response model. The code used to do this is built in such a way that a factor of 100-speed boost comes from running the computation on GPUs, a computational resource currently enjoying large-scale investment across a number of scientific disciplines.

This project seeks to enable closer collaboration between the UCL and Stockholm teams, by sending researchers on two short research stays to greatly accelerate this development process as both experimental collaborations begin to collect and analyse their first data.

Area

Physics

UCL leads

• Professor Chamkaur Ghag, Department of Physics and Astronomy