A semi‐implicit deep‐atmosphere spectral dynamical kernel using a hydrostatic‐pressure coordinate

Abstract

Most of the present deep‐atmosphere models are based on finite‐difference schemes. This article proposes a novel deep‐atmosphere non‐hydrostatic spectral dynamical kernel with hydrostatic‐pressure based terrain‐following vertical coordinate. Contrary to finite‐difference methods, a spectral transform method is used for the horizontal discretization. The prognostic variables in our dynamical kernel are analogous to the typical ALADIN‐NH spectral model. A two time level semi‐implicit semi‐Lagrangian time‐stepping scheme is applied to enable large time steps and improve the efficiency of integration. A set of test cases are conducted to evaluate the accuracy and stability of the deep‐atmosphere dynamic kernel.