Abstract
A one-dimensional (column) version of a primitive equations model has been used to study the impact of soil moisture and vegetation cover on the development of deep cumulus convection in the absence of dynamical forcing. The model includes parameterizations of radiation, turbulent exchange, deep convection, shallow boundary layer convective clouds, vegetation, and soil temperature and moisture. Multiple one-dimensional experiments were performed using the average July sounding for Topeka, Kansas, as the initial condition. A range of volumetric soil moisture from one-half of the wilting point to saturation and vegetation cover ranging from bare soil to full cover were considered.
Vegetation cover was found to promote convection, both by extraction of soil moisture and by shading the soil so that conduction of heat into the soil was reduced (thereby increasing the available energy). The larger values of initial soil moisture were found to delay the onset of precipitation and to increase the precipitation amount. The greatest rainfall amounts were generally predicted to occur for moist, fully vegetated surfaces. Vegetation cover also had a pronounced moderating influence, decreasing the sensitivity of the results to the soil moisture content. The general nature of the results prevailed for modest variations in the initial summertime atmospheric profile and changes in the details of the surface parameterization. The inclusion of shading by shallow cumulus clouds tended to reduce the convection for moist, bare (or partly bare) soil. The nonlinearity of the interaction between the land surface and convective precipitation implies that the effects of subgrid landscape heterogeneity in climate models cannot accurately be represented by linear averages of the contributions from the different surface types.
