Testing Gravity on Cosmological Scales: Theoretical Predictions with the COLA Method

TL;DR

This paper demonstrates how COLA simulations can efficiently model modified gravity theories like $f(R)$ and nDGP, enabling realistic galaxy catalogues and emulators for large-scale structure analysis.

Contribution

It introduces a method combining COLA simulations with empirical galaxy-halo models for modified gravity, and develops an emulator for the nDGP boost factor.

Findings

COLA simulations accurately reproduce large-scale structure probes in modified gravity.

The combined approach produces realistic mock galaxy catalogues in $f(R)$ and nDGP models.

An emulator for the nDGP boost factor of the matter power spectrum is successfully trained.

Abstract

One of the main objectives of stage IV galaxy surveys is to constrain gravity on cosmological scales. To this end, it is crucial to make accurate theoretical predictions in the nonlinear regime of structure formation in order to maximise the scientific return. This is possible at a relatively low computational cost thanks to COLA simulations, an approximate and much faster alternative to full $N$-body simulations. In this thesis, we focus on two modified gravity theories, $f(R)$ and nDGP, and present an analysis of how COLA simulations can be combined with empirical models for the galaxy-halo connection to produce realistic mock galaxy catalogues in modified gravity. We then use the resulting galaxy catalogues to study the effects of modified gravity and validate COLA simulations for several probes of the large-scale structure like the galaxy power spectrum, bispectrum and the void-galaxy cross-correlation function. Finally, we propose the COLA method for the extension of cosmological emulators to modified gravity theories and showcase its potential by training an emulator for the nDGP boost factor of the matter power spectrum, i.e., the ratio of the power spectrum in nDGP gravity with that in general relativity.