Advanced Numerical Methods for NWP Models

The long-term goal of this research is to construct the Navy’s next-generation global numerical weather prediction (NWP) model using new numerical methods specifically designed for distributedmemory and vector computers. To distinguish it from the current Navy global atmospheric model, the navy Operational Global Atmospheric Prediction System (NOGAPS), we shall refer to this new model as the Navy’s Spectral Element Atmospheric Model (NSEAM). To take full advantage of distributedmemory computers, the spherical global domain of NSEAM is partitioned into local sub-domains, or elements, which can then be solved independently on multiple processors. The numerical methods used on these sub-domains must be not only local in nature but also high-order accurate and highly efficient. Thus, NSEAM is being constructed so that it is as accurate as the current spectral model (NOGAPS); is more efficient, thereby allowing for finer resolution forecasts; and is geometrically more flexible, thereby allowing for the use of adaptive or telescoping grids. This will allow for better coupling with mesoscale models and eventually perhaps allowing for a modeling paradigm whereby the global and mesoscale models are either unified or at least share the same numerical algorithms which facilitates their maintenance.