Johnson et al. have recently shown that convection in the Tropics is essentially trimodal with shallow cumulus and cumulus congestus contributing to moistening and cumulus congestus and deep convection contributing to rainfall. In TOGA/COARE congestus clouds contributed to over one-quarter of the total convetive rainfall. The congestus clouds were most prevalent when there was MCS organization and cumulonimbus activity.

We have employed the Advanced Regional Prediction System model to study the formation of congestus clouds in the Tropics under disturbed periods. Convection in the model is simulated explicitly with a 6-category water/ice microphysics parameterization. Resolution of 2 km x 2 km x 500 m is used and the mean sounding from TOGA/COARE is used as a basic state sounding.

Our first goal is to study how the temperature structure produced by the stratiform component of heating in MCSs affects the developing clouds in the model. In the first set of experiments, the model has been run into a radiative convective equilibrium with periodic boundary conditions after which the two-signed temperature anomaly associated with the stratiform component of heating has been imposed. As a result the field of deep convection changes to convection which is predominantly of the congestus type. In the second set of experiments, the two-signed temperature anomaly has been imposed from the beginning of the simulation using open boundary conditions. The resulting clouds are of congestus type from the start of the simulation in this case.

We plan to study the development of congestus clouds further by, e.g., using different initial basic state soundings, imposing temperature anomalies at the melting level, and including a moisture anomaly associated with the two-signed temperature anomaly.