The Santiago-Harvard-Edinburgh-Durham void comparison I: SHEDding light on chameleon gravity tests

TL;DR

This paper compares various void finding algorithms to evaluate their effectiveness in testing chameleon f(R) gravity models, revealing that 2D voids, especially tunnel underdensities, are highly promising for future lensing surveys.

Contribution

It systematically compares void finders and introduces tunnel underdensities, demonstrating their potential to distinguish f(R) gravity from General Relativity in upcoming surveys.

Findings

Void galaxy clustering is similar across models, but dark matter density profiles differ.

2D voids, especially tunnel underdensities, can effectively test f(R) gravity.

LSST can distinguish f(R) models from GR at high significance using 2D void lensing.

Abstract

We present a systematic comparison of several existing and new void finding algorithms, focusing on their potential power to test a particular class of modified gravity models - chameleon $f(R)$ gravity. These models deviate from standard General Relativity (GR) more strongly in low-density regions and thus voids are a promising venue to test them. We use Halo Occupation Distribution (HOD) prescriptions to populate haloes with galaxies, and tune the HOD parameters such that the galaxy two-point correlation functions are the same in both f(R) and GR models. We identify both 3D voids as well as 2D underdensities in the plane-of-the-sky to find the same void abundance and void galaxy number density profiles across all models, which suggests that they do not contain much information beyond galaxy clustering. However, the underlying void dark matter density profiles are significantly different, with f(R) voids being more underdense than GR ones, which leads to f(R) voids having a larger tangential shear signal than their GR analogues. We investigate the potential of each void finder to test f(R) models with near-future lensing surveys such as EUCLID and LSST. The 2D voids have the largest power to probe f(R) gravity, with a LSST analysis of tunnel (which is a new type of 2D underdensity introduced here) lensing distinguishing at 80 and 11$\sigma$ (statistical error) f(R) models with $|f_{R0}|=10^{-5}$ and $10^{-6}$ from GR.