Currently, developing El Niño events are identified by the warming of sea surface temperatures in the equatorial Pacific Ocean. However remote clues in the atmosphere may precede the changes observed at the ocean surface. This new Prediction Index looks for signs of El Niño development in the winter, providing six to nine months lead-time.
"This statistical formula uses a time series of satellite-based data," Curtis said. "We look for rapid changes (over 30-60 day time scales) in the difference of precipitation between two specific geographic areas, namely off the coast of Sumatra (Indonesia), south of the Equator, and in the central Indian Ocean, basically on the Equator."
If this "see-saw" in precipitation is strong and the coast of Sumatra has been wetter than normal and the central Indian Ocean drier than normal over the past six months, then the formula yields a "high value," indicative of a developing El Niño. They found high values that preceded the five strongest El Niños of the study period, including the most recent 1997-1998 El Niño event.
Curtis and Adler checked satellite rainfall data on a weekly basis, and found that the index surpassed the required threshold for shifts in rainfall in late January and early February of this year. "We are therefore looking for the next El Niño to begin sometime between July and October 2002, based on the 6-9 month lag observed in the first five cases," said Adler.
"These abrupt shifts in rainfall patterns are likely accompanied by strong westerly winds that move from the Indian Ocean to the Pacific Ocean. If the winds reach the Pacific Ocean and the ocean-atmosphere conditions are right, an El Niño will develop," said Curtis.
The National Oceanic and Atmospheric Administration’s Climate Prediction Center (NOAA/CPC) is the lead agency on El Niño forecasts. Vernon Kousky of NOAA’s CPC said the new prediction tool supports past research on events leading to El Niño. "Their results are consistent with other studies that have focused on the El Niño-Southern Oscillation (ENSO)-neutral period just prior to the development of El Niños. Satellite estimates of precipitation are crucial for monitoring and predicting climate variability on intra-seasonal and inter-annual time scales," he said.
El Niño events in the tropical Pacific Ocean have impacts on global weather patterns, including increased rainfall in the eastern Pacific, and drought conditions in Indonesia and Australia.
Curtis and Adler are also using satellite information to examine these global precipitation patterns from 1979 to 2001 and how they evolve during El Niños. If the 2002-2003 event behaves like an "average" El Niño, then Curtis and Adler expect it to last about a year, with the strongest effects occurring next winter in various countries including the United States.
Curtis and Adler will use precipitation data from the NASA’s Tropical Rainfall Measuring Mission (TRMM) satellite to verify the results from their formula. TRMM, now in its fifth year of operation, has led to greatly improved rainfall information over the tropics and TRMM-based precipitation estimates will also be used to monitor the expected changes related to the El Niño.
The Global Energy and Water Cycle Experiment (GEWEX) provided the 20-year rainfall record that Curtis and Adler used to develop their Prediction Index. GEWEX is an international cooperative effort that was initiated in 1988 by the World Climate Research Program (WCRP) to observe and model the hydrologic cycle and energy fluxes in the atmosphere, and at the land and ocean surface.
Curtis presents the 2002-2003 El Niño forecast paper at the Spring American Geophysical Union meeting in Washington, D.C, at the Washington Convention Center in the H21B-04 session at 9:40 a.m., Tuesday, May 28, in Room WCC29.