A Study of the Response of Deep Tropical Clouds to Large-Scale Thermodynamic Forcings. Part II: Sensitivities to Microphysics, Radiation, and Surface Fluxes

Abstract

The Goddard Cumulus Ensemble (GCE) model is used to examine the sensitivities of multiday 2D simulations of deep tropical convection to surface fluxes, interactive radiation, and ice microphysical processes. The simulations incorporate large-scale temperature, moisture, and momentum forcings, from the Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment (TOGA COARE) for the period 19–27 December 1992. This study shows that, when surface fluxes are eliminated, the mean simulated atmosphere is much cooler and drier, convection and CAPE are much weaker, precipitation is less, and low-level to midlevel cloudiness is much greater. Surface fluxes using the TOGA COARE flux algorithm are weaker than with the aerodynamic formulation, but closer to the observed fluxes. In addition, trends similar to those noted above for the case without surface fluxes are produced for the TOGA COARE flux case, albeit to a much lesser extent. The elimination of shortwave and longwave radiation is found to have only minimal effects on the mean thermodynamics, convection, and precipitation. However, exclusion of radiation in the model does have a significant impact on cloud temperatures and structure above 200 mb. The removal of ice microphysical processes produces major changes in the structure of the clouds. Much of the liquid water is transported to the upper levels of the troposphere and evaporates, resulting in less mean total surface precipitation. The precipitation primarily occurs in regions of narrow, but intense, convective rainfall bands. The elimination of melting processes (diabatic cooling and conversions to rain) leads to greater (ice) hydrometeor mass below the 0°C level and reduced latent cooling. This, along with weaker vertical cloud mass fluxes, produces a much warmer and moister boundary layer, and a greater mean CAPE. Finally, the elimination of the graupel species has only a small impact on mean total precipitation, thermodynamics, and dynamics of the simulation, but does produce much greater snow mass just above the melting layer.