Information divergences to parametrize astrophysical uncertainties in dark matter direct detection

TL;DR

This paper introduces a method to optimize over all possible dark matter velocity distributions within a certain divergence distance from a central model, improving the robustness of direct detection limits against astrophysical uncertainties.

Contribution

It proposes a novel approach using information divergences to bracket detection limits considering uncertainties in the dark matter velocity distribution.

Findings

Brackets limits on dark matter-nucleon and dark matter-electron scattering cross sections.

Method accounts for deviations from Maxwell-Boltzmann velocity distribution.

Applicable to other physics scenarios with uncertain functions.

Abstract

Astrophysical uncertainties in dark matter direct detection experiments are typically addressed by parametrizing the velocity distribution in terms of a few uncertain parameters that vary around some central values. Here we propose a method to optimize over all velocity distributions lying within a given distance measure from a central distribution. We discretize the dark matter velocity distribution as a superposition of streams, and use a variety of information divergences to parametrize its uncertainties. With this, we bracket the limits on the dark matter-nucleon and dark matter-electron scattering cross sections, when the true dark matter velocity distribution deviates from the commonly assumed Maxwell-Boltzmann form. The methodology pursued is general and could be applied to other physics scenarios where a given physical observable depends on a function that is uncertain.