Matter power spectrum from a Lagrangian-space regularization of perturbation theory

TL;DR

This paper introduces a novel Lagrangian-space regularization method for accurately computing the matter power spectrum across linear to nonlinear scales, improving agreement with simulations.

Contribution

It proposes a new approach embedding perturbation theory within a nonlinear Lagrangian framework with an adhesion-like regularization, enhancing small-scale predictions.

Findings

Achieves <3% accuracy on large scales ($k\, extless=1 h$Mpc$^{-1}$)

Achieves <10% accuracy on small scales ($k\, extless=10 h$Mpc$^{-1}$)

Better than previous methods in matching numerical simulations

Abstract

We present a new approach to computing the matter density power spectrum, from large linear scales to small highly nonlinear scales. Instead of explicitly computing a partial series of high-order diagrams, as in perturbative resummation schemes, we embed the standard perturbation theory within a realistic nonlinear Lagrangian-space ansatz. We also point out that an "adhesion-like" regularization of the shell-crossing regime is more realistic than a "Zel'dovich-like" behavior, where particles freely escape to infinity. This provides a "cosmic web" power spectrum with good small-scale properties that provide a good matching with a halo model on mildly nonlinear scales. We obtain a good agreement with numerical simulations on large scales, better than 3% for $k\leq 1 h$Mpc$^{-1}$, and on small scales, better than 10% for $k \leq 10 h$Mpc$^{-1}$, at $z \geq 0.35$, which improves over previous methods.