The sensitivity of cloud radiative feedback to cloud representations at a global climate model

The Community Climate System Model, version 4 (CCSM4) and its super-parameterized version (SP-CCSM4) are used to study the cloud change and cloud radiative feedback to the abrupt 4x increase in the atmospheric CO2 concentration. In SP-CCSM4 cloud processes are simulated by 2D cloud resolving models embedded in each grid column of the atmospheric model, whereas CCSM4 uses the conventional parameterizations for the representation of cloud processes. The response of total cloud to surface warming shows opposite changes in the two models, and the low cloud has the largest contribution to the discrepancies. Both models simulate a negative cloud radiative effect. The cloud forcing simulated by the conventional parameterization is weaker than in the super-parameterization representation . The cloud radiative feedback in CCSM4 is larger than in SP-CCSM in which there is no significant feedback with surface warming. These large differences emphasize the fundamental uncertainties associated with conventional parameterizations in global climate model projections.