Forecasting summer/fall El Niño-southern oscillation events at 6-11 month lead times
Date
2004-06
Authors
Seseske, Stacey A., author
Journal Title
Journal ISSN
Volume Title
Abstract
Accurately predicting El Niño-Southern Oscillation (ENSO) events is an important yet challenging task, especially at the extended range of 6-11 months. This research offers a new methodology for forecasting extended range ENSO events, utilizing global data and a statistical model. Most ENSO forecasts utilize data local to the Pacific Ocean basin. This research uses an all-subsets technique to select from an inclusive pool of global predictors that are able to capture useful ENSO precursor signals beyond the Pacific basin. A multiple linear regression using the best five predictors produced a December 1 forecast for the June-July-August (6-8 month forecast) and the September-October-November (9-11 month forecast) Sea Surface Temperature Anomaly SSTA) in the Niño 3-4 region (5°N-5°S, 120-170°W). The performance of each forecast was then compared to the analogous 1 December ENSO-CLIPER (Knaff and Landsea 1911) forecast which is held as a benchmark for specifying ENSO forecast skill. Results for the 6-8 month forecast (1 December to JJA) show that the scheme presented in this research (herein referred to as SG) explains more variance than that of the ENSO-CLIPER scheme. The SG scheme accounts for 58 percent of the variance for the period 1952-2002, while the ENSO-CLIPER model explains only 14 percent for the same period. The 9-11 month forecast (1 December to SON) shows that SG only improves slightly upon ENSO-CIOPER. The SG forecast explained 3;6 percent of the variance for the 1950-2002 period, while ENSO-CLIPER explained 25 percent. However, by combining the two schemes (SG + ENSO-CLIPER) it is possible to explain over 50 percent of the variance in the SON SSTA timeseries 9-11 months in advance.
Description
June 2004.
Principal investigator: William M. Gray.
Principal investigator: William M. Gray.
Rights Access
Subject
Climatic changes -- Forecasting
El Niño Current
Southern oscillation