Regime shifts

Fig. 1. Regime shifts in (a) winter (DJF), (b) summer (JJA) and (c) annual Pacific Decadal Oscillation indices, as obtained using the regime shift detector with the following parameters: Target significance level p* = 0.1, regime length l = 20 years, and Huber parameter h = 1.

Climate regime shifts in the Pacific-North American region during the 20th century are described in detail in Rodionov (2002).  A major regime shift in the Pacific climate, from a cold to warm PDO phase, occurred in the late 1970s. There have been a lot of discussions lately that the PDO is switching back to the cold phase again (e.g, click here and here), after scientists from NASA's Jet Propulsion Laboratory announced that it was more than just a La Niña event that occurred in 2008 (see here). A similar announcement was made during the peak of a prolonged La Niña event that started in the fall of 1998 and lasted for nearly two years (see here). Soon after, however, the PDO returned to the positive territory and its winter 2003 value was the highest since 1941 (Fig. 1a). The regime shift detector applied to the winter PDO time series showed no indication of a regime shift in the late 1990s. It may be that the negative PDO value in 2008 is again just a result of the concurrent La Niña event and not a regime shift.

Fig. 2. Changes in mean winter (DJF) SST from 1978-1998 to 1999-2008.

The situation, however, is more complicated that it seems. Some argue that there was a regime shift in the late 1990s, but not of the kind that is reflected in the PDO (Bond et al., 2003). As noted by Minobe (2002), a difference between the climate shifts in the 1990s and 1970s was that the 1999 shift involved prominent warming of SSTs in the western North Pacific, whereas the cooling in this region in the 1970s was quite small.

Unlike the winter PDO index, the summer and annual PDO indices exhibit a statistically significant (p < 0.03) shift in 1999 (Fig. 1 b,c). But even in winter, the difference in SSTs between 1999-2008 and 1978-1998 closely resembles the negative PDO pattern (Fig. 2). Furthermore, the regional manifestations of the regime shift in the late 1990s are quite convincing. For example, the magnitude of the 1999 regime shift in the central North Pacific was greater that that of the 1977 shift (Fig. 3a). A highly statistically significant (p < 0.004) shift in 1999 is also evident in the waters off Baja California Peninsula (Fig. 3b).

Fig. 3. As in Fig.1, except for winter (DJF) SST in (a) central North Pacific (20-35N, 170E-160W), and (b) off Baja California Peninsula (20-25N, 125-110W).

All this indicates that 1999 marks the beginning of a new climate regime similar in many respects to a negative PDO phase (but not necessarily its classical spatial pattern). The 1999 regime shift was also found in biological records (Schwing and Moore, 2000; Batten and Welch, 2004). It is important to underscore that during a series of weak-to-moderate El Niño events in 2002/03, 2004/05, and 2006/07, SST anomalies in the central North Pacific remained positive (Fig. 3a), as during the negative PDO. The fact that the North Pacific responded stronger to La Niña than El Niño events suggests that it was a true climate regime shift in 1999.