Environmental Control of Convective Rainfall -- WTG Cloud-Resolving Model Results

Ignorance of the environmental factors that control convective rainfall and related convective properties is one of the primary stumbling blocks to more accurate weather and climate forecasts, particularly in the tropics. Field programs in which dropsondes were deployed on a dense (approximately one degree) grid in studies of tropical cyclogenesis have provided us with an extensive set of observations of tropical convection. Analysis of these results has led to measurements of vertical mass fluxes of ensembles of convection as well as the associated dynamic and thermodynamic environment. As expected, averaged vertical mass fluxes are stronger in environments with higher column relative humidities. Unexpectedly, mass fluxes are also stronger and more bottom-heavy if the moist convective (not conditional) instability is smaller. This and previous work also show a positive correlation between convection and surface heat and moisture fluxes. These results are seen in cloud-resolving model simulations of convection using weak temperature gradient (WTG) lateral boundary conditions in addition to observations.

This paper presents results of a systematic study of the intensity of precipitation and convective mass fluxes as a function of variations in environmental conditions in a WTG cloud-resolving model. In particular, a base radiative-convective equilibrium reference profile is subjected to potential temperature perturbations of order 1 K and relative humidity perturbations of order 10% with a variety of vertical structures. In addition, sea surface temperature (SST) and imposed mean surface wind speed are varied over ranges of 299-301 K and 0-20 m/s respectively. Approximately 100 simulations were made with the modified reference profiles and surface conditions. Two parameters found to be useful in characterizing the environment of observed convection, the saturation fraction (precipitable water divided by saturated precipitable water) and the instability index (1-3 km minus 5-7 km saturated moist entropy; a measure of moist convective instability) are calculated for each reference profile. A multiple linear regression of the average precipitation rate in each case is made against the four environmental variables, SST, imposed surface wind, and the reference profile saturation fraction and instability index. Together, these four variables account for over 97% of the variance in model precipitation rate, with most of this due to the SST and wind variations. To a large degree, these two variables govern the surface evaporation rate and replacing them by this single quantity in the regression produces essentially the same results.

The instability index in the reference profile correlates strongly with that in the convective domain, though the latter tends to exceed the former for small values of the reference profile instability index. There is much weaker correlation between the reference profile and convective region values of the saturation fraction, though the latter almost always exceeds the former. This presents a problem when attempting to connect WTG results with the real world; the distinction between the immediate convective environment and the distant environment (i.e., the reference profile in WTG) is less clear in the real world. Additional work is required in this area.