Testing the Accuracy of Halo Occupation Distribution Modelling using Hydrodynamic Simulations

TL;DR

This study evaluates how well halo occupation distribution models can replicate galaxy clustering in hydrodynamic simulations, revealing the importance of halo mass function corrections and additional bias parameters for accurate modeling.

Contribution

It demonstrates that with halo mass function adjustments and bias parameters, HOD models can accurately reproduce galaxy clustering from hydrodynamic simulations.

Findings

HOD models match clustering after halo mass function correction.

Baryons cause different halo mass function shifts in simulations.

Additional bias parameters improve low luminosity galaxy clustering accuracy.

Abstract

Halo models provide a simple and computationally inexpensive way to investigate the connection between galaxies and their dark matter haloes. However, these models rely on the assumption that the role of baryons can be easily parametrized in the modelling procedure. We aim to examine the ability of halo occupation distribution (HOD) modelling to reproduce the galaxy clustering found in two different hydrodynamic simulations, Illustris and EAGLE. For each simulation, we measure several galaxy clustering statistics on two different luminosity threshold samples. We then apply a simple five parameter HOD, which was fit to each simulation separately, to the corresponding dark matter only simulations, and measure the same clustering statistics. We find that the halo mass function is shifted to lower masses in the hydrodynamic simulations, resulting in a galaxy number density that is too high when an HOD is applied to the dark matter only simulation. However, the exact way in which baryons alter the mass function is remarkably different in the two simulations. After applying a correction to the halo mass function in each simulation, the HOD is able to accurately reproduce all clustering statistics for the high luminosity sample of galaxies. For the low luminosity sample, we find evidence that in addition to correcting the halo mass function, including spatial, velocity, and assembly bias parameters in the HOD is necessary to accurately reproduce clustering statistics.