Climate statistics from the National Center for Atmospheric Research community climate model CCM2

Abstract

Climatological statistics are shown from a 20‐year simulation conducted with the National Center for Atmospheric Research (NCAR) Community Climate Model, version 2 (CCM2), using an annually repeating prescribed sea surface temperature climatology. In most regards, the simulated climate is significantly improved over earlier versions of the CCM, particularly with respect to mean climatological biases. The tropospheric thermodynamic cold and dry bias that has historically plagued the CCM is largely eliminated. Additionally, several aspects of the large‐scale circulation are more faithfully reproduced, with significant improvements in the southern hemisphere circulation. Although most aspects of the CCM2 simulation are improved over those of previous model releases, there are some important elements of the climatology where the simulation is no better than, or in some cases degraded from, the earlier versions. The problem of a cold polar tropopause, although slightly improved, continues to be a major weakness of the simulation. Another more serious deficiency in the simulation is a warm bias in the northern hemisphere summer circulation, which affects the lower tropospheric and surface climate, as well as the top‐of‐atmosphere radiation budget. An aspect of the simulation that degrades certain characteristics of the midlatitude winter circulation when compared to the CCMl is an anomalous southwestward displacement of diabatic heating in the western Pacific. A large component of the northern hemisphere summer warm bias and the shift in western Pacific deep convection are related to limitations in the CCM2 diagnosis of cloud optical properties. Unrealistic nonlinear interactions between moist convection and atmospheric boundary layer processes also play a role in tropical precipitation distribution deficiencies.