The galaxy - dark matter halo connection: which galaxy properties are correlated with the host halo mass?

TL;DR

This paper explores how galaxy properties correlate with host dark matter halo mass using two models, revealing monotonic relations for some properties and complex dependencies for others, with implications for galaxy clustering analysis.

Contribution

It compares galaxy property dependencies on halo mass across two models and extends abundance matching to better reproduce clustering predictions.

Findings

Stellar and black hole masses show monotonic dependence on halo mass.

Active galactic nuclei heating affects the slope of the property-halo mass relation.

Extended abundance matching improves clustering predictions.

Abstract

We demonstrate how the properties of a galaxy depend on the mass of its host dark matter subhalo, using two independent models of galaxy formation. For the cases of stellar mass and black hole mass, the median property value displays a monotonic dependence on subhalo mass. The slope of the relation changes for subhalo masses for which heating by active galactic nuclei becomes important. The median property values are predicted to be remarkably similar for central and satellite galaxies. The two models predict considerable scatter around the median property value, though the size of the scatter is model dependent. There is only modest evolution with redshift in the median galaxy property at a fixed subhalo mass. Properties such as cold gas mass and star formation rate, however, are predicted to have a complex dependence on subhalo mass. In these cases subhalo mass is not a good indicator of the value of the galaxy property. We illustrate how the predictions in the galaxy property - subhalo mass plane differ from the assumptions made in empirical models of galaxy clustering by reconstructing the model output using a subhalo abundance matching scheme. In its simplest form, abundance matching generally does not reproduce the clustering predicted by the models, typically resulting in an overprediction of the clustering signal. We show how the basic abundance matching scheme can be extended to reproduce the model predictions more faithfully, which has implications for the analysis of galaxy clustering, particularly for low abundance samples.