Physical parameterizations for limited area models: Some current problems and issues

Summary The current opportunities of progress for short term numerical weather prediction in the domain of physical parametrizations are reviewed. Attention is paid to the fact that many models now resolve very short space and time scales, and the consequences of this situation in terms of physical parametrizations are outlined. It is argued that the most profitable areas of work are currently surface processes and cloud microphysics. The parametrization of deep cumulus convection will probably remain necessary for the next few years, but present operational schemes should be modified to take into account the breakdown of the quasi-equilibrium assumption as space and time resolutions of the models increase.With 13 Figures