Taking HOD modeling to the next level: results from SDSS and tests with hydrodynamic simulations

We model the small-scale clustering of SDSS galaxies using a greatly expanded set of clustering statistics compared to previous HOD modeling work. As a result, we achieve much tighter constraints on HOD parameters for both low- and high-luminosity galaxies. However, for our low luminosity sample, these new constraints exhibit significant tension with our previous best-fitting parameters. This tension is likely the result of incorporating new clustering statistics that are sensitive to effects like assembly bias and velocity bias, suggesting that our HOD model needs to be expanded to include these effects. We explore this further by testing the HOD modeling framework on hydrodynamic simulations. After fitting a standard HOD model to a hydrodynamic simulation and subsequently applying it to the corresponding dark matter only (DMO) simulation, we find that the HOD model is unable to reproduce the clustering of low luminosity galaxies unless the effects of spatial, velocity, and assembly bias are taken into account. Additionally, we find that the halo mass function is shifted to higher masses in the DMO simulations; as a result, the HOD model is unable to reproduce the clustering of either high or low luminosity galaxies unless a correction is made to the DMO halo masses. All of these results demonstrate the need for any future work involving HOD modeling to apply corrections to DMO halo masses, and to include parameters for spatial, velocity, and assembly bias in the HOD model.