DEMNUni: Massive neutrinos and the bispectrum of large scale structures

TL;DR

This paper presents the first measurements of the matter bispectrum in cosmologies with massive neutrinos using N-body simulations, analyzing the suppression effects and bias corrections, and comparing results with perturbation theory.

Contribution

It introduces the first measurement of the matter bispectrum in massive neutrino cosmologies from N-body simulations and evaluates perturbation theory predictions against these measurements.

Findings

Neutrino suppression on the bispectrum matches quadratic perturbation theory predictions.

Linear neutrino perturbation approximation causes errors at large scales.

Non-linear and non-local halo bias corrections are consistent with power spectrum analysis.

Abstract

The main effect of massive neutrinos on the large-scale structure consists in a few percent suppression of matter perturbations on all scales below their free-streaming scale. Such effect is of particular importance as it allows to constraint the value of the sum of neutrino masses from measurements of the galaxy power spectrum. In this work, we present the first measurements of the next higher-order correlation function, the bispectrum, from N-body simulations that include massive neutrinos as particles. This is the simplest statistics characterising the non-Gaussian properties of the matter and dark matter halos distributions. We investigate, in the first place, the suppression due to massive neutrinos on the matter bispectrum, comparing our measurements with the simplest perturbation theory predictions, finding the approximation of neutrinos contributing at quadratic order in perturbation theory to provide a good fit to the measurements in the simulations. On the other hand, as expected, a linear approximation for neutrino perturbations would lead to O($f_{\nu}$) errors on the total matter bispectrum at large scales. We then attempt an extension of previous results on the universality of linear halo bias in neutrino cosmologies, to non-linear and non-local corrections finding consistent results with the power spectrum analysis.