Modelling galaxy clustering: is new physics needed in galaxy formation models?

The clustering amplitude of galaxies depends on their intrinsic luminosity. We compare the properties of publicly available galaxy formation models with clustering measurements from the Two-degree Field Galaxy Redshift Survey. The model predictions show the same qualitative behaviour as the data but fail to match the observations at the level of accuracy at which current measurements can be made. We demonstrate that this is due to the model producing too many satellite galaxies in massive haloes. We implement simple models to describe two new processes: satellite-satellite mergers and the tidal dissolution of satellites to investigate their impact on the predicted clustering. We find that both processes need to be included in order to produce a model which matches the observations.