Monte Carlo Simulation Variance Reduction Techniques for Photon Transport in Liquid Xenon Detectors

TL;DR

This paper presents an accelerated Monte Carlo simulation method for photon transport in liquid xenon detectors, significantly reducing computation time while maintaining accuracy, which is crucial for background analysis in large-scale dark matter experiments.

Contribution

The paper introduces a novel variance reduction technique that efficiently simulates photon interactions within LXe detectors, enabling faster background modeling for large-scale experiments.

Findings

Achieved three orders of magnitude speed-up in simulations.

Validated the method for gamma-induced background modeling.

Applicable to large-scale liquid xenon detector simulations.

Abstract

Monte Carlo simulations are a crucial tool for the analysis and prediction of various background components in liquid xenon (LXe) detectors. With improving shielding in new experiments, the simulation of external backgrounds, such as induced by gamma rays from detector materials, gets more computationally expensive. We introduce and validate an accelerated Monte Carlo simulation technique for photon transport in liquid xenon detectors. The method simulates photon-induced interactions within a defined geometry and energy range with high statistics while interactions outside of the region of interest are not simulated directly but are taken into account by means of probability weights. For a simulation of gamma induced backgrounds in an exemplary detector geometry we achieve a three orders of magnitude acceleration compared to a standard simulation of a current ton-scale LXe dark matter experiment.