Improving and understanding convection and cloud simulation in the NCAR GCM using ARM observations

Description

In an effort to improve the representation of convective processes in the NCAR GCM, we made several modifications to the convection scheme in the NCAR CAM3 and CAM5. These include closure, convection trigger, convective momentum transport, and convective cloud microphysics. These modifications significantly improved the model simulations of the mean climate and variability such as Madden-Julian Oscillation (MJO). To understand these improvements, we analyzed the heat and moisture budgets for MJO. It is found that lower troposphere moistening prior to deep convection, which is important for MJO initiation, is a combined result of vertical advection and shallow convection. Without shallow convection, the moistening is much slower. For the interaction between convection and large-scale clouds, the hydrometeor budgets are computed in the tropical and midlatitude convective regions. We found that different microphysical processes play different roles in convective and large-scale clouds. Cloud microphysical properties from the model are compared with ARM and other observations. In addition, with the inclusion of convective microphysics in the model, we will also explore the interaction of aerosol and convection using CAM5.