Observable tests of self-interacting dark matter in galaxy clusters: cosmological simulations with SIDM and baryons

TL;DR

This study uses large-scale cosmological simulations including both self-interacting dark matter (SIDM) and baryons to identify observable differences from cold dark matter, especially in gravitational lensing, aiding future dark matter model constraints.

Contribution

First large-volume cosmological simulations with SIDM and baryons, analyzing their effects on galaxy cluster observables and lensing to differentiate dark matter models.

Findings

SIDM haloes are less dense and rounder than CDM haloes.

Baryons do not significantly alter SIDM halo density profiles.

Gravitational lensing can distinguish SIDM from CDM, with constraints on SIDM cross-section.

Abstract

We present BAHAMAS-SIDM, the first large-volume, (400/h Mpc)^3, cosmological simulations including both self-interacting dark matter (SIDM) and baryonic physics. These simulations are important for two primary reasons: 1) they include the effects of baryons on the dark matter distribution 2) the baryon particles can be used to make mock observables that can be compared directly with observations. As is well known, SIDM haloes are systematically less dense in their centres, and rounder, than CDM haloes. Here we find that that these changes are not reflected in the distribution of gas or stars within galaxy clusters, or in their X-ray luminosities. However, gravitational lensing observables can discriminate between DM models, and we present a menu of tests that future surveys could use to measure the SIDM interaction strength. We ray-trace our simulated galaxy clusters to produce strong lensing maps. Including baryons boosts the lensing strength of clusters that produce no critical curves in SIDM-only simulations. Comparing the Einstein radii of our simulated clusters with those observed in the CLASH survey, we find that at velocities around 1000 km/s an SIDM cross-section of sigma/m > 1 cm^2/g is likely incompatible with observed cluster lensing.