The Southern Oscillation and Tree Rings: 1600–1961

View More View Less
  • 1 Laboratory of Tree-Ring Research, University of Arizona, Tucson, AZ 85721
© Get Permissions Rent on DeepDyve
Restricted access

Abstract

Variations of the Southern Oscillation (SO) significantly influence surface climate over a large part of the globe. In the first part of this study we show that teleconnection patterns known from the instrumental record to be associated with the SO also appear to be present in both dependent and independent tree-ring reconstructions of climate during the 19th and 20th centuries. Reconstructions of surface temperature and precipitation appear to be more closely related to the SO than those of pressure. Since these dendroclimatic reconstructions appear to be responsive to climate anomalies related to the SO, it should be possible to reconstruct past values of the SO directly from tree-ring width variations. Thus, we attempted to reconstruct an index of the SO back to 1600 A.D. using tree-ring chronologies from both western North America and the Southern Hemisphere. The best estimates were derived from the western North American chronologies alone. These estimates calibrate about 50 percent of the SO variance with about 20 percent of the variance verified over an independent period. Significant spectral peaks are apparent at periods in the two to ten year range of the SO, Low-index events are estimated to be less frequent during the 19th century than before or since and extreme events in 1792–93 and 1815–16 may have been of comparable magnitude to the event of 1982–83.

Abstract

Variations of the Southern Oscillation (SO) significantly influence surface climate over a large part of the globe. In the first part of this study we show that teleconnection patterns known from the instrumental record to be associated with the SO also appear to be present in both dependent and independent tree-ring reconstructions of climate during the 19th and 20th centuries. Reconstructions of surface temperature and precipitation appear to be more closely related to the SO than those of pressure. Since these dendroclimatic reconstructions appear to be responsive to climate anomalies related to the SO, it should be possible to reconstruct past values of the SO directly from tree-ring width variations. Thus, we attempted to reconstruct an index of the SO back to 1600 A.D. using tree-ring chronologies from both western North America and the Southern Hemisphere. The best estimates were derived from the western North American chronologies alone. These estimates calibrate about 50 percent of the SO variance with about 20 percent of the variance verified over an independent period. Significant spectral peaks are apparent at periods in the two to ten year range of the SO, Low-index events are estimated to be less frequent during the 19th century than before or since and extreme events in 1792–93 and 1815–16 may have been of comparable magnitude to the event of 1982–83.

Save