Relationships between Atmospheric Internal Variability and the Responses to an Extratropical SST Anomaly

The January and February responses of a GCM to an imposed extratropical Pacific SST anomaly are compared with the patterns of the model's internal variability. Relevant patterns of internal variability are diagnosed from ensembles of model control runs by regressing monthly mean geopotentials and temperatures against low-level temperatures in the vicinity of the SST anomaly and by EOF analysis. These patterns are found to play a significant role in determining the local and the remote responses to the SST anomaly. Different responses to the SST anomaly in the GCM's January and February climates are largely explained by the differences in the regression patterns and in the leading EOF. The GCM response may be considered as comprising a direct linear response to low-level heating, that is local to the forcing and baroclinic, and an eddy-forced component that closely resembles patterns of the model's internal variability—it is equivalent barotropic and extends over the entire hemisphere. The results suggest that for a warm SST anomaly over the Kuroshio Extension to induce an equivalent-barotropic ridge immediately east of the anomaly, the internal variability must have a well-defined center of action over the central Pacific. In this GCM, this is nearly true in February but not in January. Similar analyses are performed for the observed flow to determine the patterns of variability in nature and thereby to suggest the potential response to SST forcing. The natural variability in January and February has a strong large-scale center over the Pacific, which, according to the model results, should favor the development of an equivalent-barotropic ridge in response to a warm SST anomaly.