http://arxiv.org/abs/1702.02620
We present a clustering comparison of 12 galaxy formation models (including Semi-Analytic Models (SAMs) and Halo Occupation Distribution (HOD) models) all run on halo catalogues and merger trees extracted from a single {\Lambda}CDM N-body simulation. We compare the results of the measurements of the mean halo occupation numbers, the radial distribution of galaxies in haloes and the 2-Point Correlation Functions (2PCF). We also study the implications of the different treatments of orphan (galaxies not assigned to any dark matter subhalo) and non-orphan galaxies in these measurements. Our main result is that the galaxy formation models generally agree in their clustering predictions but they disagree significantly between HOD and SAMs for the galaxies which are not assigned to any subhalo. The scatter between the models on the 2PCF when orphan satellites are included can be larger than a factor of 2 for scales smaller than 1 Mpc/h. We also show that galaxy formation models that do not include orphan satellite galaxies have a significantly lower 2PCF on small scales, consistent with previous studies. Finally, we show that the distribution of orphan satellites within their host halo is remarkably different between SAMs and HOD models. While HOD models distribute orphan satellites using a theoretical (usually a NFW) profile independently of the merger trees, the positions and trajectories of orphan satellites in SAMs are analytically evolved based on the position and trajectory of the subhalo they belonged to initially. Because of this, orphan satellites in SAMs are more correlated and present a higher small-scale clustering than in HOD models. We conclude that orphan satellites have an important role on galaxy clustering and they are the main cause of the differences in the clustering between HOD models and SAMs.
A. Pujol, R. Skibba, E. Gaztanaga, et. al.
Fri, 10 Feb 17
19/46
Comments: 17 pages, 7 figures
You must be logged in to post a comment.