Consideration on computational advantages of a high-order dynamical core based on the discontinuous Galerkin method in atmospheric turbulent simulations
Consideration on computational advantages of a high-order dynamical core based on the discontinuous Galerkin method in atmospheric turbulent simulations