An improved representation of PBL moist processes for large-scale numerical models will be presented with results produced by the UCLA Atmospheric General Circulation Model (AGCM) and coupled atmosphere-ocean GCM (CGCM).
The earlier version of the UCLA coupled atmosphere-ocean GCM (CGCM) produced a cold bias in the simulated mean sea surface temperature(SST) and a very weak interannual variability. Experiments with the CGCM show that simulated SST is extremely sensitive to formulations of physical processes in the GCM, especially in its AGCM component. It has been found that, in these experiments, surface evaporation and marine stratus cloud incidence play significant roles in determining SST.
Earlier experiments with the old version of the model indicate that the surface evaporation rates are too high, the subtropical marine stratus are poorly simulated, humidity in the layer immediately above the PBL is too high. These suggest that the formulations of the PBL moist processes in the AGCM need to be revised.
We have made many revisions in the AGCM including: 1) the formulation of the properties of air entraining into the PBL, 2) the process after the stratus cloud becomes unstable, 3) the upgrading of the cloud shortwave radiative properties and 4) the effect of subgrid-scale topographic variations on PBL stratus cloudiness. The results of simulated seasonal mean stratus incidence with these revisions indicate a much more realistic stratocumulus incidence over subtropical marine stratocumulus regimes. In addition, the model also simulates precipitation, surface evaporation, sea level pressure, surface wind stress, shortwave and longwave surface fluxes, net surface heat fluxes, and large-scaleatmospheric circulations far more realistically than the simulations performed with our earlier version.
Some of the results will be discussed in the presentation.