Variability of 500-MB Geopotential Heights in a General Circulation Model and the Projection of Regional Greenhouse Effect Climate Change.

Many researchers have utilized general circulation models (GCMs) in establishing climate change scenarios for specific regions or locations, despite the mismatch of spatial scales involved. A major underlying assumption involved in utilizing model output in this manner is that the GCM contains mid-tropospheric dynamics that are internally consistent with those of the real climate system. The main purpose of this study is examine the forms and processes of mid-tropospheric variability in the Goddard Institute for Space Studies (GISS) GCM, with the hope of shedding light on this model-analog strategy. The response of mean 500 mb and surface air temperature fields in the GISS GCM to a doubling of CO_2 indicates a substantial relationship between the two. Unfortunately, the GISS GCM demonstrates systematic flaws in its simulation of mid-tropospheric dynamics. These are revealed in an examination of high-frequency and low -frequency 500-mb teleconnections in the model. The shapes and amplitudes of known teleconnection patterns are not well simulated. This is likely due to the weak stationary wave structure found in the control run of the model. More importantly, several model teleconnections appear to coincide geographically with the patterns of mean climate change. This may indicate a direct relationship between the modelled mid-tropospheric dynamics and the spatial patterns of mean climate change. This finding has two important implications. First, it is necessary to further study the influence of GCM mid -tropospheric dynamics on the spatial distribution of climate changes being modelled. Second, and more fundamentally, spatially specific climate system feedbacks may be substantially affected by variations in teleconnection strength and frequency, potentially impacting the global climate far beyond the regional scale.