Hurricane forecasts not blowing us away

MIAMI -- In the 2005 Atlantic hurricane season, the worst in recorded history, the most famous of hurricane forecasters was way off the mark when he predicted there would be 15 tropical storms. There were 28.

In fact, the past three seasons were a bust for the Colorado State University prediction team founded by Bill Gray, the pioneer of long-range hurricane forecasts, and others who dare to foray into the unpredictable world of cyclones.

The failures have left storm researchers and emergency managers asking themselves if there's any point to the seasonal forecasts that get so much attention in the media and are important to the 89 million people in the U.S. hurricane danger zone and tens of millions more in the Caribbean, Mexico and Central America.

"I wish people wouldn't hang their hat on these numbers," said Tom Iovino, a communications specialist with Florida's Pinellas County.

Iovino and others who communicate storm dangers to the public say their work is complicated by the seasonal outlooks issued by Gray, the U.S. government, private forecaster AccuWeather, London-based Tropical Storm Risk and others.

Seasonal forecasts are a best guess at the big picture. For example, a forecaster might warn it's going to be a busy season -- the entire Atlantic basin and the Caribbean could see 15 tropical storms, and nine of them could become hurricanes.

But the chance of a hurricane hitting any one particular area is actually very small. And as weather specialists are fond of saying -- if only one hits you, it's a bad season.

Emergency managers use the pre-season forecasts to promote preparedness for the June 1-Nov. 30 season. It's a way to tell people to stockpile food and water, pack an emergency kit with flashlights and fresh batteries and plan an evacuation route.

But if predictions for a busy season are wrong, Iovino says people inevitably grumble about having prepared for nothing.

"They'll say, 'Well, what happened to those predictions?' " he said. "In a tough economy, some people ask, 'Why am I wasting my money?' "

Worse, he said, they might not pay attention next year, or when a storm is actually threatening them.

The U.S. National Oceanic and Atmospheric Administration is going to downplay its forecast this year and will emphasize instead whether the season will be above or below the long-range averages, rather than the actual number of storms.

An average season has about 10 tropical storms, of which about six become hurricanes.

"The seasonal forecast is meaningless to preparedness," U.S. National Hurricane Center director Bill Read said.

But the forecasts are eagerly awaited in financial markets, in part because of the blockbuster hurricane seasons of 2004 and 2005, which caught everyone off guard.

Energy markets trembled every time a storm rolled into the oil- and gas-producing Gulf of Mexico, and insurance markets blanched at the damage caused by four 2004 storms that hit Florida, and 2005's all-time costliest hurricane, Katrina.

The last time Gray's CSU team was close to the mark was 2004, when it predicted 14 storms and there were 15. In 2006 it foresaw 17 storms and there were only 10, and in 2007 it projected 17 and there were 14.

Gary's team on Wednesday raised the number of tropical storms and hurricanes it expects to form in the upcoming Atlantic storm season.

It increased its outlook by two tropical storms to 15, and by one hurricane to eight, compared with a long-term average of around 10 and six, respectively, for a storm season.

"Current oceanic and atmospheric trends indicate that we will likely have an active Atlantic basin hurricane season," Gray said.

Of the eight hurricanes predicted by the forecasters for the six-month season starting June 1, four were forecast to become major storms with winds of at least 111 miles per hour . Major, or intense, storms, which rank from Category 3 to Category 5 on the five-step Saffir-Simpson scale of hurricane intensity, are the most destructive.

Researchers say hurricane forecasting is as unpredictable as the weather.

They must stir in all the unstable ingredients that could affect a hurricane -- sea temperatures in the Atlantic, upper-level winds, the El Nino and La Nina conditions in the eastern Pacific, rainfall in Africa, even dust over the ocean.

Gray, who issued his first hurricane forecast in 1984, notes that the predictions were never intended to be seen as absolute. They only discuss probabilities.

"Many people don't understand probabilities," he said.

"I suppose there are some negatives with it, but overall you want the best information out there," Gray said.

(Editing by Michael Christie and Philip Barbara)

REUTERS