One of the most fundamental properties of the global heat balance is the net heat input into the tropical atmosphere that helps drive the atmospheric circulation across all space and time scales. Among the distinguishing aspects of the net heating are the following: (i) the varying spatial/temporal structure is a combination of net heating and cooling components, (ii) the tropical atmosphere continually loses energy radiatively to space while taking energy from the land / ocean surface, and (iii) the dynamical response to this varying heat source / sink forcing is a transport of energy sufficient to render the climate at any location essentially non-local. Although much is understood in broad terms about the gross structure of the heating, the physical processes involved and even interannual variations, our ability to express this understanding in predictive models is still rather poor.

The work reported here will explore processes that determine tropical diabatic heating associated with the global monsoons and focus on differences between several global climate models and steadily improving satellite retrievals over the past decade. We examine the nature of convective processes as part of the energy transfer mechanism between the land and ocean surface (experiencing net radiative heat input) and the radiatively cooled atmosphere. We are conducting several numerical experiments in which the closure of model convective and cloud parameterizations, and the nature of surface properties is altered. These experiments are designed to isolate the sensitivity of precipitation efficiency and cloud production to interactions between the convective closure and surface fluxes. Statistics of boundary layer CAPE and its modulation by combined surface forcing and convection is one way in which we examine the surface / PBL / free atmosphere energy exchange important to determining atmospheric diabatic heating. To interpret the results of these experiments, a number of satellite retrievals are used: Top-of-atmosphere radiative fluxes from ERBS and CERES are used to examine shortwave and longwave cloud forcing and to deduce required seasonal energy transports over land areas. Retrievals of cloud properties from ISCCP, TOVS Pathfinder A, MODIS and AMSU are also used to relate the cloud structure to precipitation variations. TRMM retrievals of reflectivity, rainfall rate, and inferred diabatic heating are analyzed to show both seasonal and intraseasonal variations in vertical structure of latent heat release.

With these experiments and observational measurements we aim to discern not only how the model simulations of diabatic heating depart from reality, but also the extent to which model convective and cloud processes behave in a manner physically and statistically consistent with observations. Our results also highlight how these processes differ among heating centers over the Amazon, central Africa, the Austral-Asian region, and the western Pacific Warm Pool.