On Parametrization of Convective and Nonconvective Latent Heat Release

Abstract

Convective adjustment procedures remove conditional instability in a vertical sounding preserving the total energy (latent, internal and potential). It is shown that this procedure has very undesirable properties in the very first time step in numerical weather prediction, e.g., large-scale temperature and moisture distributions are greatly altered. If, on the other hand, convective adjustment is carried out on a mesoscale, as is the case for the Kuo parameterization procedure, then the large-scale conditional instability is preserved and the changes in initial data are small in the first (and subsequent) time steps. The latter procedure is used to evaluate convective precipitation in the vicinity of a squall line. A detailed heating function is designed for use in the ω equation and primitive equation prediction models. This function permits stable or unstable heating to occur in a given region and also allows for the simultaneous occurrence of both kinds of heating in the same region. Precipitation rate estimates show major improvement over earlier versions which followed only stable heating.