A Cloud Model to Parameterize Convection

Abstract

This paper describes a parameterization of convection developed for use with a multi-level primitive equation model of the atmosphere. This parameterization is similar in form to that of Kuo, who postulates that boundary layer air rises moist adiabatically before mixing with the large-scale flow. A major difference from Kuo arises, however, when we allow condensate (the cloud material) to be evaporated to the large-scale flow during the mixing process. This parameterization conserves moisture and energy of the large-scale flow in a fashion similar to Manabe, and is suitable for general circulation studies as well as short-term prediction. The parameterization tends to adjust the vertical profile of temperature to the pseudo-adiabatic lapse rate from the boundary layer, and the relative humidity tends to saturation. Adjustments are made frequently to points in the tropics. Results are compared with a similar integration using Manabe's convective adjustment scheme. The parameterization presented maintains stable integration, and has certain advantages when compared to the Manabe parameterization. The new parameter introduced, the maximum amount of cloud material at any level, can be obtained from cloud model results and may also be measured directly.