Non-universality of the mass function: dependence on the growth rate and power spectrum shape

TL;DR

This study demonstrates that the halo mass function depends on growth history and power spectrum shape, challenging the assumption of universality, and introduces a parameterization that improves predictive accuracy across diverse cosmologies.

Contribution

The paper provides evidence of non-universality in the halo mass function and introduces a new parameterization accounting for growth rate and power spectrum shape effects.

Findings

Mass function depends on growth history and power spectrum shape.

New parameterization accurately models halo abundance across cosmologies.

Improved cosmology-rescaling methods achieve 2% accuracy.

Abstract

The abundance of dark matter haloes is one of the key probes of the growth of structure and expansion history of the Universe. Theoretical predictions for this quantity usually assume that, when expressed in a certain form, it depends only on the mass variance of the linear density field. However, cosmological simulations have revealed that this assumption breaks, leading to 10-20% systematic effects. In this paper we employ a specially-designed suite of simulations to further investigate this problem. Specifically, we carry out cosmological N-body simulations where we systematically vary growth history at a fixed linear density field, or vary the power spectrum shape at a fixed growth history. We show that the halo mass function generically depends on these quantities, thus showing a clear signal of non-universality. Most of this effect can be traced back to the way in which the same linear fluctuation grows differently into the nonlinear regime depending on details of its assembly history. With these results, we propose a parameterization with explicit dependence on the linear growth rate and power spectrum shape. Using an independent suite of simulations, we show that this fitting function accurately captures the mass function of haloes over cosmologies spanning a vast parameter space, including massive neutrinos and dynamical dark energy. Finally, we employ this tool to improve the accuracy of so-called cosmology-rescaling methods and show they can deliver 2% accurate predictions for the halo mass function over the whole range of currently viable cosmologies.