A Dark Census: Statistically Detecting the Satellite Populations of Distant Galaxies

TL;DR

This paper develops a statistical framework to detect and analyze the population of dark matter subhalos around distant galaxies through their collective gravitational lensing effects, extending beyond the strong lensing region.

Contribution

It introduces a general formalism for statistically detecting dark satellite populations via lensing observables, applicable to various halo geometries and subhalo mass functions.

Findings

Lensing observables follow Gaussian distributions in the presence of subhalos.

The formalism reduces computational costs for modeling substructure effects.

Enables Bayesian inference for satellite population detection.

Abstract

In the standard structure formation scenario based on the cold dark matter paradigm, galactic halos are predicted to contain a large population of dark matter subhalos. While the most massive members of the subhalo population can appear as luminous satellites and be detected in optical surveys, establishing the existence of the low mass and mostly dark subhalos has proven to be a daunting task. Galaxy-scale strong gravitational lenses have been successfully used to study mass substructures lying close to lensed images of bright background sources. However, in typical galaxy-scale lenses, the strong lensing region only covers a small projected area of the lens's dark matter halo, implying that the vast majority of subhalos cannot be directly detected in lensing observations. In this paper, we point out that this large population of dark satellites can collectively affect gravitational lensing observables, hence possibly allowing their statistical detection. Focusing on the region of the galactic halo outside the strong lensing area, we compute from first principles the statistical properties of perturbations to the gravitational time delay and position of lensed images in the presence of a mass substructure population. We find that in the standard cosmological scenario, the statistics of these lensing observables are well approximated by Gaussian distributions. The formalism developed as part of this calculation is very general and can be applied to any halo geometry and choice of subhalo mass function. Our results significantly reduce the computational cost of including a large substructure population in lens models and enable the use of Bayesian inference techniques to detect and characterize the distributed satellite population of distant lens galaxies.