Statistical Method Developed to Predict Pacific Sea Surface Temperatures


February 27, 2007

Michael Alexander, Cecile Penland, Luda Matrosova and James Scott, researchers at the NOAA Earth System Research Laboratories (ESRL) and the University of Colorado's Cooperative Institute for Research in the Environmental Sciences (CIRES) in Boulder, CO, have developed a statistical method to predict seasonal sea surface temperatures (SSTs) over the tropical and north Pacific Ocean. The model is also used to forecast the Pacific Decadal Oscillation (PDO), the time series indicating the magnitude and sign of the leading climate pattern of North Pacific SST anomalies. To date general circulation models have had very limited success in predicting SSTs in midlatitudes. The statistical forecasts exhibit significant skill over much of the Pacific, for two-to-three seasons in advance and up to a year in some locations, particularly for forecasts initialized in winter. The predicted PDO is significantly correlated with observations for forecasts of up to one year in advance.

Background:
The statistical method, termed linear inverse modeling (LIM), has proved to be successful in predicting El Ni–o conditions in the tropical Pacific. Here LIM has been extended to include the north Pacific, where both numerical and statistical models have had limited success in forecasting SSTs. The ESRLĐCIRES research team focused on predicting the PDO as it is an important index associated with widespread changes in climate and ecosystems over the Pacific/North American region. The LIM-based PDO forecasts are more skillful than other statistical methods, such as persistence or autoregressive models, and have comparable skill to LIM forecasts of SSTA associated with El Niño.

Significance:
This experimental tool shows significant promise as a viable means to aid in forecasting climate events at longer lead-times and may be especially useful for fishery managers and hydrologists. This research supports NOAA's mission goal of understanding climate variability and change to enhance society's need to plan and respond.

Contact: Michael Alexander More Information: