GCM (general circulation model)-data intercomparison: The good news and the bad

General circulation models (GCMs) are being actively used to assess possible climate change due to increasing greenhouse gas concentrations. Because such simulations provide detailed climatic predictions at a wide range of scales, they are of particular interest to those making regional assessments of climatic change. It is especially important that workers using the results of such simulations be aware of some of the limitations of these results. In this study some of the positive results from these model simulations will be shown and some of the deficiencies will also be highlighted. Following an introductory section describing the nature of GCM climate simulations the issue of the spatial scales of such simulations is examined. A comparison of the results of seven GCM simulations of the current climate and the predictions of these models for the changes due to a doubling of CO{sub 2} will be discussed. In these intercomparisons, the spatial scale over which the results are compared varies from global to zonal (longitudinally averaged at a given latitude) to individual slices through the data along specified latitudes or longitudes. Finally, the dangers and pitfalls of relying on simple averages will be highlighted. 19 refs., 9 figs., 1 tab.