Predictions for the SKA from hierarchical galaxy formation models

There is now overwhelming evidence to suggest that structure in the Universe formed hierarchically. The development of collapsed structures in the dark matter due to gravitational instability has been studied extensively using numerical simulations and analytic techniques. Modelling the baryonic component of the Universe is much more challenging. The process of galaxy formation and evolution can be followed in the context of hierarchical structure formation using the technique of semi-analytical modelling. Due to our lack of knowledge of the pertinent physics, some parts of the model are more uncertain than others, one clear example being the parameterization of the timescale for star formation. The SKA will help to remove this ambiguity by testing predictions for the evolution of the distribution of star formation rates and the neutral hydrogen mass function. Here we give examples of such predictions for two models which differ in how the star formation timescale in galaxies scales with redshift. We use the GALFORM semi-analytical model to compute the evolution of the neutral hydrogen mass function with redshift. The star formation histories generated by the GALFORM model are combined with the spectro-photometric code GRASIL to generate spectral energy distributions for model galaxies that extend beyond the optical to sub-millimetre and radio wavelengths. We use this hybrid code to make predictions for the number counts of radio sources whose radio emission is driven by star formation.