Simulated Ocean–Atmosphere Interaction in the North Pacific from a GCM Coupled to a Constant-Depth Mixed Layer

Abstract

The life cycle and structure of dominant wintertime SST anomalies and associated atmospheric response in the extratropical North Pacific are examined using results from a 100-yr seasonal simulation of a low-resolution atmospheric model with realistic geography coupled to a simple mixed layer ocean. The study focuses on composited SST anomalies produced solely by ocean?atmosphere energy exchange. One key pattern shows a negative (positive) SST anomaly in the central Pacific, denoted CCP (WCP), flanked by opposite signed anomalies in the western and eastern Pacific. For the WCP case, the SST anomaly reaches about 1.0°C in the central Pacific, whereas for the CCP case it is ?1.5°C. During the growth phase of the WCP (CCP) SST anomalies, anomalous highs (lows) occur over the western Pacific and over western North America, and an anomalous low (high) is over the east-central Pacific. To the rear of the anomalous low (high), a negative (positive) SST anomaly develops in response to anomalous cold, dry (warm, moist) air. The effect of anomalous wind also contributes but to a lesser extent. The composited SST anomalies primarily develop in 1?2 months. During the decay stage of the WCP SST anomaly, the atmospheric anomalies are essentially of opposite sign than during the growth stage and help to destroy the SST anomaly. In contrast, the atmospheric anomalies during the decay stage of the CCP SST anomaly are of the same sign but weaker than during the growth stage. For this case, a positive SST-atmosphere feedback involving ocean?atmosphere energy exchange helps maintain a more persistent SST anomaly than in the WCP composite case. Comparison of results with prescribed SST anomaly experiments indicate that the ocean is in part forcing the atmosphere during the decay stage of CCP SST anomalies. The magnitude and position of the lower-tropospheric anomalies forced by SST anomalies are similar to those obtained in linear theory, while the vertical structure is more akin to that found in higher-resolution GCMs with realistic geography. The weaker atmospheric anomalies generated by the low-resolution model are likely a consequence of its simulation of weak transients. Differences between the composite and prescribed SST anomaly results (e.g., in position of the atmospheric anomalies relative to the SST anomaly) suggest that the anomalous response of the atmosphere to SST anomalies in a coupled ocean?atmosphere environment may not be the same as in a one-way forced system.