The Pacific Decadal Oscillation (PDO) is a North Pacific Ocean climate state that varies on a multi-decadal time scale between two modes with distinct spatial and temporal SST characteristics. The SSTs of central interest correspond to a large area in the northern and western Pacific Ocean and a smaller region in the eastern tropical Pacific (Mantua et al., 1997). Since PDO is identified on the basis of SSTs and is associated with periodicities ranging from 20 to 30 years, it has been described as a long-lived El Nino-like pattern of Pacific climate variability. The PDO is in a "warm or positive phase'' when temperatures are anomalously warm in the eastern tropical Pacific and anomalously cool in the central North Pacific. Warm PDO phases dominated from 1925 to 1946 and from 1977 to at least 1998 (Hare and Mantua, 2000). A "cool or negative'' PDO phase has anomalously cool temperatures in the eastern tropical Pacific and anomalously warm temperatures in the central North Pacific. The PDO was in a cool phase from 1890 to 1924 and from 1947 to 1976 (Mantua et al., 1997). Although the PDO is much slower to switch from one phase to another than the fast-fluctuating ENSO, SSTs characterizing the PDO undergo fairly abrupt changes from one phase to another termed regime shifts (Miller et al., 1994). Accumulating evidence indicates an important climate regime shift occurred in the North Pacific Ocean in the winter of 1976-7 (Trenberth, 1990: Ebbesmeyer et al., 1991; Zhang et al., 1997). Another regime shift may have occurred in 1998 coincident with the end of the 1997-8 El Nino and the beginning of the subsequent La Nina event (Hare and Mantua, 2000). Verdon and Franks (2006) study of paleo-records
1900 1920 1940 1960 1980 2000 Date
Fig. 8.9. The Pacific Decadal Oscillation (PDO) index for January 1900 to July 2006.
The gray line is the monthly value, and the bold line is a 7-month moving average.
(Data courtesy NOAA and the University of Washington Joint Institute for the Study of
Atmosphere and Ocean from their website at http://jisao.washington.edu/pdo/.)
suggests that a relationship between ENSO frequency and PDO phase changes has existed for the last 16 phase changes, and Biondi et al. (2001) and Hidalgo (2004) conclude that PDO and ENSO interactions have occurred throughout the last four centuries.
The PDO is portrayed by an index derived from the leading principal component from an un-rotated normalized empirical orthogonal function analysis of gridded mean November through March SSTs poleward of 20° N latitude in the Pacific Basin (Mantua et al., 1997). Positive PDO index values identify a warm phase and indicate months of above normal SSTs along the west coast of North and Central America and along the equator (Fig. 8.9). Below normal SSTs are present in the central and western North Pacific Ocean at about the latitude of Japan. Negative PDO values indicate a cool phase and are associated with below normal SSTs along the equator and above normal SSTs east of Japan (Trenberth and Hurrell, 1994). Although the PDO shares some SST features with ENSO, the PDO is set apart from ENSO by two main characteristics. Typical PDO events show greater persistence than ENSO events, and evidence of the PDO is most visible in the North Pacific and North America. The ENSO signature features an opposite pattern of SSTs with stronger evidence in the topics (Mantua et al., 1997; Hare and Mantua, 2000). The possibility that the appearance and strength of ENSO events may depend on how PDO dominates ocean circulation and temperature patterns is receiving increasing attention.
The physical mechanisms responsible for the PDO have not been identified, but several possible mechanisms involving SSTs, mean oceanic currents and
Rossby waves, and atmospheric forcing have been proposed (Miller and Schneider, 2000). Identifying the precise mechanisms responsible for PDO variations is complicated by the interaction of feedback processes involving ocean-atmosphere and tropical-extratropical linkages (Trenberth and Hurrell, 1994). This is illustrated by the reported intensification of North Pacific winter cyclones since 1948 that spans the most recent PDO cool and warm phases (Graham and Diaz, 2001). An additional factor complicating efforts to understand PDO mechanisms is the suggestion that the PDO is characterized by two general periodicities of from 15 to 25 years and from 50 to 75 years (Chao et al., 2000). However, PDO indices from the 1600s reconstructed from tree rings indicate an average regime duration of 23 years (Biondi et al., 2001). Even in the absence of theoretical understanding, empirical PDO climate information is useful for season-to-season and year-to-year climate forecasts for the Pacific, Australia, and the Americas because of its strong tendency for multiseason and multiyear persistence. Also, proxy sources have yielded a clear pattern of symmetric atmospheric circulation changes associated with the PDO that signal an influence of this phenomenon on climate in the tropics and the Southern Hemisphere (Linsley et al., 2000).
Over the past century, the amplitude of the PDO climate pattern has varied irregularly at interannual-to-interdecadal time scales (see Fig. 8.9). The difference in SSTs from positive to negative phases is not more than 1 to 2 °C, but the affected ocean area is so huge that it can impact North American temperature and precipitation patterns through its effect on steering storms across the North Pacific Ocean (Mantua et al., 1997). PDO warm phases produce above average winter and spring temperatures in northwestern North America and below average temperatures in the southeastern United States. The southern United States and northern Mexico receive above average winter and spring rainfall during warm PDO phases, but precipitation is below average in the interior Pacific Northwest and the Great Lakes region. Cool phase PDO temperatures and precipitation are broadly the reverse of these climate anomaly patterns across North America (Mantua et al., 1997; Nigam et al., 1999). Severe and sustained droughts in the western United States over the past 200 years show a relationship to SSTs indexed by the PDO (Gedalof et al., 2004; Hidalgo, 2004), and streamflow variability in the United States displays a significant correlation with PDO warm and cool phases (Tootle and Piechota, 2006). McCabe and Dettinger (2002) found that the 1 April snowpack in the western United States is more highly correlated with PDO than the Nino-3 SST, but Stewart et al. (2005) conclude that evidence is lacking to support warmer temperatures related to PDO as the major contributor to a one-to-four week advance in peak streamflow in snowmelt dominated watersheds across a broad region of the western United States.
In general terms, North American climate anomalies associated with PDO warm and cool phases are broadly similar to El Niño and La Niña patterns. However, closer inspection reveals that positive PDO phases enhance El Niño conditions and weaken La Niña effects while negative PDO phases weaken El Niño events and enhance La Niña events (Gershunov and Barnett, 1998; McCabe and Dettinger, 1999). Consequently, the coincidence of the PDO phase and ENSO events produces a complex array of wet and dry conditions that complicates generalizations addressing the regional wetness and dryness associated with El Niño and La Niña events. Emerging evidence indicates that the commonly portrayed El Niño and La Niña temperature and precipitation patterns are only valid during years in which a warm PDO coincides with an El Niño event and a cool PDO coincides with a La Niña event (Gershunov and Barnett, 1998; Nigam et al., 1999).
Was this article helpful?