A Lagrangian Perturbation Theory in the presence of massive neutrinos

TL;DR

This paper develops a Lagrangian Perturbation Theory framework to accurately model the clustering of cold dark matter in cosmologies with massive neutrinos, incorporating non-linear effects and biasing to match simulations.

Contribution

It introduces a third-order LPT approach that accounts for neutrino effects and ensures well-behaved large-scale behavior, improving modeling of matter and tracer correlation functions.

Findings

Accurate correlation functions down to 20 Mpc/h at z=0.5

Good agreement with Quijote simulations without free parameters

Effective modeling of redshift-space distortions and tracer biasing

Abstract

We develop a Lagrangian Perturbation Theory (LPT) framework to study the clustering of cold dark matter (CDM) in cosmologies with massive neutrinos. We follow the trajectories of CDM particles with Lagrangian displacements fields up to third order in perturbation theory. Once the neutrinos become non-relativistic, their density fluctuations are modeled as being proportional to the CDM density fluctuations, with a scale-dependent proportionality factor. This yields a gravitational back-reaction that introduces additional scales to the linear growth function, which is accounted for in the higher order LPT kernels. Through non-linear mappings from Eulerian to Lagrangian frames, we ensure that our theory has a well behaved large scale behavior free of unwanted UV divergences, which are common when neutrino and CDM densities are not treated on an equal footing, and in resummation schemes that manifestly break Galilean invariance. We use our theory to construct correlation functions for both the underlying matter field, as well as for biased tracers using Convolution-LPT. Redshift-space distortions effects are modeled using the Gaussian Streaming Model. When comparing our analytical results to simulated data from the Quijote simulation suite, we find good accuracy down to $r=20 \,\text{Mpc} \, h^{-1}$ at redshift $z=0.5$, for the real space and redshift space monopole particle correlation functions with no free parameters. The same accuracy is reached for the redshift space quadrupole if we additionally consider an effective field theory parameter that shifts the pairwise velocity dispersion. For modeling the correlation functions of tracers we adopt a simple Lagrangian biasing scheme with only density and curvature operators, which we find sufficient to reach down to $r=20 \,\text{Mpc} \, h^{-1}$ when comparing to simulated halos.