The Effects of Cumulus Moisture Transports on the Simulation of Climate with a General Circulation Model

Abstract

A hybrid cumulus parameterization was developed to study the effects of cumulus heat and moisture transports on climate simulations. This parameterization consists of a cumulus mass flux representation of the convective fluxes of heat and moisture, and moist adiabatic adjustment representation of the release of latent heat and the removal of moisture by precipitation processes. This hybrid scheme was implemented and tested in the National Center for Atmospheric Research Community Climate Model. Averages over the last 90 days of 210-day simulations made for January and July with the standard moist-convective adjustment scheme were compared with those obtained with the hybrid scheme. The simulations made with the hybrid parameterization resulted in upper-tropospheric temperatures (p ∼ 300 mb) that were as much as 8°C higher than the standard scheme in the tropics. In addition, the tropospheric water content simulated with the hybrid parameterization was considerably greater than that of the standard model. Both the temperature and water vapor simulations obtained with the hybrid scheme were in much closer agreement with the observations than those of the standard model. In addition, the hybrid scheme resulted in more realistic simulations of precipitation distributions and the upper-tropospheric flow patterns in the tropics. Convective activity associated with deep and shallow convection was in qualitative agreement with observed distributions. In the tropics, the convective heating associated with the hybrid parameterization occurs higher in the troposphere than that of the standard scheme. Although the differences between the zonal mean convective heating profiles obtained with the two parameterizations are large, these differences are largely compensated for by changes in the zonal mean radiative heating due to changes in temperature, humidity, and cloudiness. Thus our results suggest that the zonal mean tropical troposphere, when perturbed, undergoes a sort of radiative-convective adjustment. This result, if borne out by more detailed model studies, would lead to the important conclusion that the cumulus parameterization problem is intricately coupled to the parameterization of cloud-radiative interactions.