Model intercomparison study: cloud-radiative forcings and feedbacks

Effects of cloud-radiation schemes on cloud forcing and feedback are tested using two different GCMs (general circulation models): the NCAR (National Center for Atmospheric Research) CCM2 (Community Climate Model Version 2) and YONU (Yonsei University GCM). The major differences between the cloud-radiation schemes are in the method of treating cloud water content: in CCM2 the cloud water content is prescribed, while in YONU the cloud water content is computed explicitly. Perpetual July integrations driven by globally constant sea surface temperature forcings of two degrees, both positive and negative, are employed. The total net cloud forcing is negative in both models, i.e., cloud cools the Earth-atmosphere system. The net cloud-radiative feedback of CCM2 is negative, i.e., cloud moderates the global warming, while that of YONU is positive, i.e., cloud amplifies it at both the top of the atmosphere and at the surface. These consist of quite different shortwave and longwave components between the two models. We have validated the resulting cloud-radiative forcings of the present climate by using ERBE (Earth Radiation Budget Experiment) and ISCCP (International Satellite Cloud Climatology Project) datasets. As judged by these observations, the resulting fields generated by these two models are qualitatively right, i.e., both simulated results show the right sign of forcing, but quantitatively wrong in many aspects, especially in the shortwave components. In order to simulate the future climate reasonably, there needs much effort to improve the parameterizations of cloud-radiative properties in the climate models.