Sensitivity of Diagnosed Convective Fluxes to Model Assumptions

Abstract

The sensitivity of diagnosed cloud mass and heat fluxes over tropical oceans to assumptions about cumulus-scale updrafts and downdrafts is tested. The ensemble of clouds investigated is the population of precipitating clouds observed in GATE. The basic input to the calculations is radar data. However, a previous paper shows that the same sensitivity is involved in diagnostic calculations based on large-scale heat and moisture budgets. The assumptions investigated include the relationship between entrainment rate and cloud size, thermodynamic conditions at the base of convective updrafts. thermodynamic properties of entrained air, thermodynamic conditions at the tops of convective downdrafts. and the mass flux profiles in the updrafts and downdrafts. Diagnostic results are most sensitive to the relation between cloud size and entrainment rate and the thermodynamic conditions at the base of the updraft. It is found that assumed mass flux profiles that do not concentrate all of the detrainment from updrafts at cloud top lead to apparently more reasonable heating profiles in the high troposphere. Generally, however, the uncertainty in the diagnostic results that can be attributed to cumulus-scale model assumptions is small, no greater than the uncertainty inherent in the basic data.