Hurricane Forecaster a Winner
BOSTON — A STORM warning is already flying for next year's North Atlantic hurricane season.
Forecaster William Gray predicts activity "somewhat above" the average of the last 42 years. He expects 11 storms strong enough to receive names. Six of them should become hurricanes, including three severe hurricanes, he says.
Dr. Gray is the atmospheric scientist at Colorado State University at Fort Collins who correctly predicted last fall the level of this year's Atlantic hurricane activity. He predicted four hurricanes, including one severe storm - Hurricane Andrew.
This year's hurricane season, which ended Nov. 30, was relatively mild. Only six storms were named, including the four hurricanes. Gray notes, however, that "people impacted by Andrew will have a hard time thinking 1992 was an inactive year." The season also was unusual in that no hurricane formed in the tropics. All of them, including Andrew, formed above 25 degrees north latitude.
"I was lucky," Gray says, in this year's hurricane forecast. He adds that he hopes "people will remember this year when I don't do so well in the future." But his general forecasting scheme has shown this kind of so-called "luck" for seven of the past eight years. It failed in 1989.
What is remarkable now is the long lead time of these forecasts. Gray used to issue his initial prediction just as the official Atlantic hurricane season opened June 1. He then updated the forecast as the most intense part of the season began Aug. 1. Now he issues his initial forecast in late November for the following year's hurricane season.
Meteorologist Rainer Bleck at the University of Miami at Coral Gables, Fla., says, "Many people [in the atmospheric science community] find [Gray] credible and admire him for his detective work." He explains that Gray does not claim too much for his scheme. He doesn't try to forecast when and where specific storms may form or their course. Professor Bleck says this restraint lends credibility to Gray's work.
Gray and his colleagues have identified large-scale weather factors that can be assessed in November and that account for 45 to 50 percent of the variance among hurricane seasons. Gray calls this a "surprising" finding. "I know of no other [long-range weather] prediction scheme with this much skill," he says.
The forecasts are based on close observation of the large-scale weather factors. These observations yield numbers that Gray plugs into a statistical formula that, in turn, yields the hurricane prediction.
These factors include the presence and strength of the widespread warming of tropical Pacific Ocean surface water called El Nino. A strong El Nino suppresses Atlantic hurricane activity. The factors also include the strength and direction of stratospheric winds that blow around the globe above the equator. These winds reverse direction about every two years. West African rainfall is another predictor. Gray says that failure to take proper account of this rainfall ruined his 1989 forecast.
The forecast can give insurance companies, relief agencies, and hurricane monitors a hint of how severe a season may be. Gray notes that before his work, such planners had nothing to go on except climatological averages.