July simulation by the UCLA general circulation model

Abstract

Results of a July simulation produced by the UCLA general circulation model are analyzed with a view to improving the model. It is shown that while many features of the July climatology are well reproduced, others, such as the intensity of the jet streams in the upper troposphere, the frequency of cyclogenesis, the structure of the subtropical pressure belt in the Southern Hemisphere, and the precipitation are not well reproduced. It is suggested that the sources of error are related to inadequate vertical and horizontal resolutions, to the choice of upper boundary at 50 mb, to poor representation of topographically forced motions, and to the indirect coupling between the general circulation model and the planetary boundary model. The analysis has led to the introduction of major design changes in the model.