The Earth's radiation budget and its relation to atmospheric hydrology: 3. Comparison of observations over the oceans with a GCM

This paper reports on a study that seeks to examine a very limited set of interactions between the hydrological cycle and the radiative processes that take place on Earth and attempt to test how well these interactions are simulated by a general circulation model. Two broad types of tests of the model are introduced in this paper. The first focuses on comparing various sensitivity relationships established by the model with those observed on Earth. The specific relationships examined in this paper include the sensitivities of column‐integrated water vapor, the clear sky greenhouse effect, cloud albedo, and cloud radiative forcing to sea surface temperature. The second type of test focuses on comparison of simulated and observed seasonal cycles of the greenhouse effect and cloud radiative forcing. The main results of this study suggest that the model studied, which we take to be representative of the current generation of global climate models, is able to simulate some aspects of the observed sensitivities fairly well. Qualitative successes of the simulations include realistic variations of column vapor with sea surface temperature, the clear sky greenhouse parameter and its variation with both column vapor and sea surface temperature, and cloud longwave radiative forcing with sea surface temperature, as well as the seasonal changes of the greenhouse effect and the cloud radiative forcing. However, there exist many serious quantitative differences between the model and the observations. These include problems in the simulations of the column vapor in the tropics and an excessively strong clear‐sky greenhouse effect in the mid‐latitudes. These differences combine in such a way that the model underestimates the sensitivity of the clear‐sky greenhouse to changes in sea surface temperature compared to that derived from observations. Significant differences between the simulated and the observed radiative properties of clouds include the overprediction of the longwave cloud radiative forcing over warm tropical oceans and a significant underestimate of the cloud reflection in mid‐latitudes during the summer season.