Abstract
The spatial structure and temporal characteristics of prominent anomalies occurring in a 15-year simulation with a GFDL spectral general circulation model are examined using empirical orthogonal functions, teleconnection patterns, composite charts, lagged correlation functions and frequency spectra.
Despite the absence of any nonseasonal perturbation in the prescribed forcing such as sea surface temperature, insolation and cloud cover, the simulated circulation exhibits an appreciable degree of temporal variability on monthly time scales. The standing oscillation in the Northern Hemisphere winter which accounts for the largest fraction of this variance has a coherent three-dimensional structure. In the middle and upper troposphere, this preferred mode of oscillation is characterized by a wavelike pattern with multiple centers of action. The corresponding anomaly pattern at the sea level is dominated by north–south pressure seesaws over the North Atlantic and North Pacific. The flow patterns associated with these pressure anomalies are consistent with the principal temperature anomaly pattern in the lower troposphere. The large-scale features of the above anomaly patterns are similar to those associated with the most prevalent standing oscillation observed in the atmosphere. The synoptic behavior and hydrological processes in the model atmosphere during the outstanding anomalous episodes are internally consistent.
The spatial structure of the principal mode in the simulation is rather insensitive to the averaging period of the model data. The autocorrelation function and frequency spectrum of the first principal component, as determined from daily data, are characteristic of persistent phenomena with no preferred periodicity. The autocorrelation time scale associated with this anomaly pattern is estimated to be ∼ 15 days.
The principal anomaly pattern in the Northern Hemisphere summer is relatively less organized, while those for the Southern Hemisphere and the tropics are noted for their zonal symmetry. The east–west sea level pressure seesaw associated with the observed Southern Oscillation over the Pacific is not simulated by the model, thus suggesting the potential role of nonseasonal forcing mechanisms (such as sea surface temperature anomalies) in that phenomenon.
