Parameterized Slantwise Convection in a Numerical Model

Abstract

A convective parameterization that mimics the theoretical effects of slantwise convection initiated by the release of conditional symmetric instability (CSI) is tested in a three-dimensional primitive equation model. Great positive impact is noted. The scheme is simplest to consider as a Kuo-type convective scheme that operates along (slanted) momentum surfaces. Because of the longer time scales associated with slantwise convection, the slantwise parameterization initiates only in vertical columns where the normal upright convective scheme does not. The slantwise parameterization activates frequently in the area where CSI has been suggested to have been released on the date tested. This is in an area of strong convection observed in an area that is stably stratified in the vertical. It is also the most unstable area as measured by thermodynamic variables along momentum surfaces. Other aspects of the parameterization are in line with theoretical expectations. When compared (early in the forecast period) to runs with only an upright convective scheme, the runs including the slantwise parameterization produce less grid-resolved and more convectively generated (meaning slantwise and upright) precipitation. In addition, the grid-resolved precipitation generated is not as organized into lines. Atmospheric structures similar to those observed near fronts are produced more readily by the model runs including the slantwise parameterization. Differences between runs with and without the slantwise parameterization at 12 and 18 h suggest that the slantwise convective parameterization may be helping the model overcome the spinup problem by more quickly generating divergent circulations. Finally, much greater precipitation amounts (far more in line with reality) are produced by the runs including slantwise convection.