indications of prior shifts as well (Minobe, 1997). The suite of atmospheric and oceanic changes that have been linked to these basin-wide regime shifts is collectively referred to as the Pacific Decadal Oscillation (PDO) (Mantua et al., 1999).
The sea surface temperature (SST) patterns associated with the PDO and ENSO are similar, the main distinction being that the extratropical features are somewhat more prominent in the PDO pattern. As in the few-year variations associated with the swings between El Niño (warm) and La Niña (cold) conditions in the equatorial Pacific, warm and wet decades in the equatorial zone tend to be marked by extratropical circulation patterns that favor an unusually active storm track in the mid-Pacific that splits toward its eastern end. An unusually large fraction of disturbances moves northeastward, bringing mild, wet weather to the Alaska panhandle; many of the remainder track southeastward, bringing heavy rains to southern California and the US desert Southwest. The mountain ranges of British Columbia and the US Pacific Northwest, which lie directly downstream from the split in the storm track, tend to receive less than the normal amount of winter snowfall, and this reduces water supplies for the following summer season. The dynamic mechanisms responsible for the long-range “teleconnections” between the equatorial Pacific and the extratropics are better understood than the processes that control the evolution of this phenomenon on the decadal time scale. Hence, regime shifts such as the one that occurred in 1976-1977 are difficult to diagnose in real time, let alone to predict.
There are several different schools of thought as to the nature of the interdecadal PDO variability, which has shown both the abruptness and persistence to qualify under our definition of abrupt climate change. The default hypothesis is that the PDO is merely a reflection of stochastic variability originating in the atmosphere but amplified by positive feedbacks associated with coupling between the atmosphere and ocean (Bretherton and Battisti, 2000). If this interpretation is valid, it follows that this ENSO-like variability is inherently unpredictable (i.e., that it becomes clearly evident only with the benefit of hindsight). Hopes that the phenomenon is deterministic, and therefore predictable, are based on the notion that ocean dynamics play an active role in PDO evolution, to the extent of setting the time scale for the major swings back and forth between the positive and negative polarity of the PDO pattern. One oceanic process that could conceivably set the time scale is the recirculation time for water parcels in the clockwise North Pacific and counterclockwise South Pacific subtropical