A Next Generation Atmospheric Prediction System for the Navy

Abstract : A long-term goal of this project is to develop a global cloud-permitting (3-km resolution) forecast capability as part of the multi-agency next-generation Earth System Prediction Capability (ESPC) initiative. Within the next decade or less, computational platforms with more than 1,000,000 processors will be commonly available and high-fidelity global weather forecasts with variable resolution grids at the mesoscale or cloud permitting scales will be routinely feasible. Unfortunately, current generation weather models are incapable of performing such forecasts because of their lack of variable resolution grid capabilities, inadequate numerical accuracy, and computational inefficiency. A paradigm shift in the numerical weather and climate prediction capabilities of the US Navy and the nation is needed in order to be at the leading edge in the future. The time is right and critical to embark on an accelerated development path of a modeling system with next-generation attributes since a decade of development is typically required to advance a Numerical Weather and Climate Prediction (NWCP) system to the point where it is operationally viable. Global and regional models are currently applied as separate modeling systems at many operational centers such as the Navy, AFWA, NCEP, with the global models providing lateral boundary conditions for the local models with finer spatial resolution over areas of interest. Recently, the horizontal spatial resolution of global models used in the leading centers around the world is fast approaching the grid spacing traditionally used by the mesoscale models, which is on the order of 10 km. In the relatively near future, when the resolution becomes even finer, approaching cloud permitting scales, the Nations global model dynamical cores will need to account for non-hydrostatic effects. A new class of next-generation models is emerging.