Please choose your delivery country and your customer group
The status of research and development on numerical weather prediction at a new (smaller) scale is reviewed. This regional scale uses a horizontal mesh size of about 35 kM over a horizontal domain size of about 1700 kM squared. The model forecasts weather, including clouds and stratiform (steady) and cumulus (showery) precipitation out to 24 h beyond the latest observation time. Technological developments permitted this research to begin in 1970 and a prototype system for operational weather forecasting will be completed in 1980. The principal gains are expected to be in the accuracy and detail of precipitation forecasts and cloudiness forecasts, including the effects of terrain and (parameterized) cumulus convection. The principal problem area is the development of techniques to incorporate into the preceding forecast, more recent, detailed observations from surface-based and satellite observations. Other problems include parameterization of small scale physical processes, particularly those associated with cloudiness and topography, such that these processes properly interact with the dynamics of scales resolved by the model. Statistical treatment of model output, expression of predictions in statistical terms, and meaningful verification measures are all very important to an effective forecast system. The optimization of the entire system is important to both the quality and the cost of an operational system. (ERA citation 03:029308)