ECMWF's Integrated Forecast System (IFS) shares key model components with Meteo-France, among others a non-hydrostatic option for applications at horizontal scales finer than about 10 km. At hydrostatic scales (for grid-sizes up to 10 km) the non-hydrostatic IFS gives very similar forecasts to the operational hydrostatic IFS, and can be run stably with rather long time steps that are used with the latter model. However, the computational cost of the non-hydrostatic IFS per time step is substantially larger than with the hydrostatic IFS (about twice at 10 km resolution).
With the PantaRhei project (FP7/2012/ERC Advanced Grant agreement no. 320375) ECMWF invests in a longer-term development of a new dynamical core that fulfils the requirement to operate at global scales, across a wide range of resolutions with utmost stability and accuracy, and with reasonable computational cost. The project explores a hybrid approach for forecasting global weather and climate that combines the strengths of established and efficient (at large hydrostatic scales) structured grid numerical weather prediction and climate models with control-volume (small-scale) edge-based codes, originating from other computational fluid dynamics disciplines.
The presentation will present status and plans of model development, and also show other efforts to enhance the efficiency of the model.