The National Center for Atmospheric Research community climate model version 2 (CCM2) simulation of the circumpolar trough, surface air temperature, the polar vortex, cloudiness, winds, and atmospheric moisture and energy budgets are examined to validate the model's representation of the present-day Antarctic climate. The results show that the CCM2 can well simulate many important features over Antarctica, such as location and intensity of the circumpolar trough, the coreless winter over the plateau, the intensity and horizontal distribution of the surface inversion, the speed and streamline pattern of the katabatic winds, the double jet stream feature over the southern Indian and Pacific oceans, and the arid climate over the continent. However, there are also some serious errors in the model. Some are due to old problems but some are caused by the new parameterizations in the model. The model errors over high southern latitudes can be summarized as follows: The circumpolar trough, the polar vortex, and the westerlies in midlatitudes are too strong; the semiannual cycle of the circumpolar trough is distorted compared to the observations; the low centers of the circumpolar trough and the troughs in the middle and upper troposphere are shifted eastward by 15¿--40¿ longitude; the surface temperatures are too cold over the plateau in summer and over the coastline in winter; the polar tropopause continues to have a cold bias; and the cloudiness is too high over the continent.

These biases are induced by two major factors: (1) the cloud optical properties in tropical and middle latitudes, which cause the eastward shift of trough and surface low centers and the error in the semiannual cycle, and (2) the cold bias on the surface air temperature, which is attributed to the oversimulation of cloudiness over the continent, especially during summer, and the uniform 2-m-thick sea ice. The constant thickness of sea ice suppresses the energy flux from the ocean to the atmosphere and hence reduces the air temperature near the coast during winter. Finally, although the simulated Antarctic climate still suffers these biases, the overall performance of the CCM2 is much better than that of the CCM1-T42. Therefore, CCM2 is good enough to be used for climate change studies, especially over Antarctica. ¿ American Geophysical Union 1994