Predicting storm's strength, path is a tricky task

Hurricane forecasters refine their techniques while hedging their bets

September 03, 2004|By Frank D. Roylance | Frank D. Roylance,SUN STAFF

Still unable to say precisely where Hurricane Frances will strike tomorrow, the National Hurricane Center hedged its bets yesterday and raised hurricane warning flags along 300 miles of Florida's east coast.

"They want to be on the safe side," said Hans Graber, director of the University of Miami's Center for Southeastern Tropical Advanced Remote Sensing.

With sustained winds of 145 mph at its core and hurricane-force winds spanning an area the size of West Virginia, Frances is about to rip into a long stretch of vulnerable - and valuable - real estate.

"It's a huge storm, and diameter-wise you could basically put the entire length of Florida in there," he said. The hurricane-force winds alone reach 80 miles outward from the eye.

Hurricane warnings were posted yesterday from Florida City, south of Miami, to Flagler Beach, north of Daytona.

People and property

Millions live between those two points - far more than were affected last month by Hurricane Charley. And they sit on some of the nation's most valuable property. Just south of the storm's forecast path, such tony havens as Palm Beach, Boca Raton and Fort Lauderdale were bracing today for winds of at least 74 mph, dangerously high water, or both.

And it could be much worse, Graber said: "It's moving now over very warm water in the Bahamas and, it's slowing down due to the blockage of a high-pressure system in the north. And this is, of course, ripe for the storm to intensify."

The slowing of the storm's forward speed also makes it more unpredictable - just as Charley made life miserable for forecasters and Gulf Coast residents who did not know exactly where the storm would strike until the last moment.

Graber compared a slow-moving storm to an old-fashioned weather vane: "When the [wind] speed becomes very low, the vane flops around in all directions. It's the same situation with a hurricane. If it really slows down to a speed where there's very little forward movement, it's very difficult to say ... what it will do."

Frances, moving at just 13 mph last night, was expected to slow even more by today.

Forecasters at the National Hurricane Center acknowledged that their computer models, while improving, are still limited in their ability to pinpoint the storm's track. "There's quite a bit of uncertainty," said Lt. Dave Roberts, a Navy meteorologist.

Margin of error

In fact, the map showing a storm's predicted track includes a feature unself-consciously called an "error cone." The official forecast track is flanked by a large blob shaped more like a light bulb than a cone. It encompasses all the other possible tracks, to the left or right, that weather computers have proposed, based on an analysis of past forecasting errors.

Hurricane forecasters watch the storm's real-time behavior and compare it to many computer predictions. The models that seem to be doing the worst job - in hindsight - are re-aligned with reality and run again.

"With time, the models become more in line with each other," Roberts said.

Even as late as yesterday, the official track showed low pressure in the Midwest drawing Frances more to the northwest - crossing the state to Tallahassee and on into Alabama. But some models still had the storm veering north into South Carolina, or across Florida into the Gulf of Mexico.

As much as numerical prediction has improved over the past 20 years, Roberts said, small errors can make a big difference.

"With Charley, we were only 40 miles off on our track," Roberts said, "but that was significant to our public." Many residents of southwestern Florida were angry when the storm crashed ashore well south of where most had expected it.

"We have to convey to the public not to put too much weight on the forecast track," he said, "but [rather ] on the extent of damage that could happen as it moves onshore."

A slow-moving hurricane is not only hard to predict, but it also lingers longer, doing more damage over time and piling up ocean water into even bigger, deadlier storm surges.

On Wall Street yesterday, Risk Management Solutions, which does disaster modeling for insurance companies, predicted that Frances would cost the industry at least $10 billion in insured losses, based on the forecast track. That would make it the second-costliest natural disaster in U.S. history, after Hurricane Andrew in 1992.

Miami's Graber recalled that he and his family were huddled in their home near Miami when Hurricane Andrew came ashore in 1992.

"Andrew was moving at 20 to 25 mph," he said. "It was just barreling down on us. Where I was, it started at 4 a.m., and by 8 o'clock it was over.

"For about four hours it was like I had an infinite freight train running by my house."

His family survived, but their windows blew out and their house was shot through by a neighbor's roof tiles, fired at them by Andrew at 140 mph.

Four hours was bad enough, he said. "But if you have double that time, or more, you can do much more damage."

Baltimore Sun Articles
Please note the green-lined linked article text has been applied commercially without any involvement from our newsroom editors, reporters or any other editorial staff.