A coarse grained perturbation theory for the Large Scale Structure, with cosmology and time independence in the UV

TL;DR

This paper introduces a coarse-grained perturbation theory for the Universe's large-scale structure that combines perturbative methods with N-body simulations, achieving high accuracy in predicting the nonlinear power spectrum across various cosmologies.

Contribution

It develops a hybrid approach that accurately reproduces the nonlinear power spectrum using a simple 1-loop IR computation and a single correlator from simulations, applicable across different cosmologies without additional simulations.

Findings

Achieves percent-level accuracy in power spectrum predictions up to k~0.4 h/Mpc.

Demonstrates that the correlator measured in one cosmology can be rescaled for others.

Shows the approach's relation to Effective Field Theory methods.

Abstract

Standard cosmological perturbation theory (SPT) for the Large Scale Structure (LSS) of the Universe fails at small scales (UV) due to strong nonlinearities and to multistreaming effects. In Pietroni et al. 2011 a new framework was proposed in which the large scales (IR) are treated perturbatively while the information on the UV, mainly small scale velocity dispersion, is obtained by nonlinear methods like N-body simulations. Here we develop this approach, showing that it is possible to reproduce the fully nonlinear power spectrum (PS) by combining a simple (and fast) 1-loop computation for the IR scales and the measurement of a single, dominant, correlator from N-body simulations for the UV ones. We measure this correlator for a suite of seven different cosmologies, and we show that its inclusion in our perturbation scheme reproduces the fully non-linear PS with percent level accuracy, for wave numbers up to $k\sim 0.4\, h~{\rm Mpc^{-1}}$ down to $z=0$. We then show that, once this correlator has been measured in a given cosmology, there is no need to run a new simulation for a different cosmology in the suite. Indeed, by rescaling this correlator by a proper function computable in SPT, the reconstruction procedure works also for the other cosmologies and for all redshifts, with comparable accuracy. Finally, we clarify the relation of this approach to the Effective Field Theory methods recently proposed in the LSS context.