Resolving the problem of galaxy clustering on small scales: any new physics needed?

Abstract

Galaxy clustering sets strong constraints on the physics governing galaxy formation and evolution. However, most current models fail to reproduce the clustering of low-mass galaxies on small scales ($r<1Mpc/h$). In this paper we study the galaxy clusterings predicted from a few semi-analytical models. We firstly compare two Munich versions, Guo et al. (2011, Guo11) and De Lucia \& Blazoit (2007, DLB07). The Guo11 model well reproduces the galaxy stellar mass function, but over-predicts the clustering of low-mass galaxies on small scales. The DLB07 model provides a better fit to the clustering on small scales, but over-predicts the stellar mass function. These seem to be puzzling. We find that there is slightly more fraction of satellite galaxies residing in massive haloes in the Guo11 model, which is the dominant contribution to the clustering discrepancy between the two models. However, both models still over-predict the clustering at $0.1Mpc/h<r<10Mpc/h$ for low mass galaxies. This is because both models over-predict the number of satellites by $30\%$ in massive halos than the data. Actually, the better agreement of DLB07 model with the data on small scales comes as a coincidence as it predicts too many low-mass central galaxies which are less clustered and thus bring down the total clustering. Finally, we show the predictions from the semi-analytical of Kang et al. (2012). We find that this model can simultaneously fit the stellar mass function and galaxy clustering if the supernova feedback in satellite galaxies is stronger. We conclude that semi-analytical models are now able to solve the small-scales clustering problem, without invoking of any other new physics or changing the dark matter properties, such as the recent favored warm dark matter.