Observational Support
Monthly averaged ERBE observations of cloud longwave radiative forcing (CRF) during the 1985-89 period, and Reynolds' analyses of SST during the same period, can each be broken down into four components. Those SST components which average to zero over a large area are associated with large shifts of CRF toward the higher SSTs, of 20-25 W/m2/K. These include the annual component of the seasonal cycle and the time average distribution within the Pacific warm pool region. Conversely, components involving mean SST changes over large areas (areas that include most deep convective activity) are not associated with significant changes in CRF. These components include El Niñ o, La Niñ a, and the biannual component of the seasonal cycle (see Figure 16.1). This does not prove that overall cloudiness is insensitive to SST, however, since it is influenced also by non-local events such as the Asian monsoon.
Model Support
A simple box model of the tropics is presented which is based on energy and moisture conservation, and simple dynamics. The model has two locations, representing convective and non-convective regions, and two layers (Figure 16.2). It is forced by SST as a boundary condition. This model fails to show a large change in cloud cover when the SST in either region is changed. However, when the model is reduced in size to cover two halves of the convecting region only, it shows sensitivity to SST redistribution within this region on the same order as the observed sensitivity to zero-mean SST shifts. The lack of sensitivity of the total cloud cover is tentatively explained as a result of the energy conservation requirement in the upper layer, assuming a direct proportionality between CRF and convective heating.
The model results (still preliminary) suggest that cloud shortwave forcing by deep clouds is the dominant negative feedback damping SST heterogeneities within convecting regions, but surface fluxes could be the dominant mechanism in stabilizing overall SST changes throughout convective regions. Neither of these mechanisms is very effective in stabilizing the model to tropical mean SST changes. In reality, cloud optical properties or lifetimes (which are constant in the model) might allow clouds to provide feedbacks for these overall perturbations.
References
Zhang, C., 1993: Large-scale variability of atmospheric deep convection in relation to sea surface temperatures in the tropics, J. Climate, 6, 1898-1913.
Hartmann, D.L., and M.L. Michelson, 1993: Large-scale effects on the regulation of tropical sea surface temperatures, J. Climate, 6, 2049-2062.
