The discontinuous Galerkin method is used to solve the non-linear spherical shallow water equations with Coriolis force. The numerical method is well-balanced and takes wetting/drying into account. The two fold goal of this work is a comparative study of dynamic and static tsunami generation by seabed displacement and the careful validation of these source models. The numerical results show that the impact of the choice of seabed displacement model can be significant and that using a static approach may result in inaccurate results. For the validation of the studies, we consider measurements from satellites and buoy networks for the 2011 Tohoku event and the 2004 Sumatra-Andaman tsunami. The results confirm that the method is appropriate for large-scale tsunami simulations and early warning systems.