A Potential Predictability Study Conducted with an Atmospheric General Circulation Model

Abstract

This paper describes a potential predictability study on the results of a 20.5 year simulation conducted with the Canadian Climate Centre (CCC) General Circulation Model (GCM). The CCC GCM is an atmosphere GCM with surface hydrology, soil moisture and snow cover representations. Except for these terms, the only source of interannual variability in the CCC GCM-simulated climate is ?internal? dynamics. The study addresses the question of whether an atmospheric GCM of this sort can simulate interannual variability which is potentially predictable in a statistical sense. Strong evidence is found for potential predictability of 500 mb height and surface pressure in the December, January, February (DJF) season of the simulated climate, and somewhat weaker evidence in the March, April, May (MAM) season. There is little evidence of potential predictability during the other seasons. The evidence does not indicate that potential predictability is related to the presence of surface hydrology, soil moisture and snow cover terms in the CCC GCM. The results have the same general characteristics as those which have been obtained in atmospheric studies: F-ratios measuring potential predictability are generally smallest in midlatitudes and larger in the tropics and towards the poles. Also, consistent with atmospheric studies, F-ratios are larger in the Southern Hemisphere than in the Northern Hemisphere, and are largest in the DJF season. However, unlike the atmosphere, it is impossible to link potential predictability in the simulated climate with any sort of quasi-periodic phenomenon such as the Quasi-biennial Oscillation or the El Niño?Southern Oscillation. Potential predictability in the simulated climate appears to arise from the occurrence of a single large anomaly extending over a period of about a season, during which atmospheric mass is systematically shifted from the tropics to the high latitudes of the Southern Hemisphere. Such large extended anomalies have previously been observed in other climate simulations and in the atmosphere.