Challenging storms

Sharper forecasts, better warnings


Hurricane's Path Of Destruction

September 19, 2004|By Michael Hill | Michael Hill,SUN STAFF

It is one thing to build a structure that can resist a hurricane. It is quite another to figure out if that structure is going to be hit by one.

Hurricane forecasting has come a long way since the days when these storms were rumors reported by those from the islands or ships at sea. Decades ago, coastal residents could do little more than nervously stare out into the ocean with wonder, knowing that danger lurked somewhere in the unknown.

Now, these storms' fearful symmetry is spectacularly evident on satellite photographs, their positions plotted mile by mile. Those satellites probe them with a variety of instruments, while aircraft regularly fly into and around them to retrieve a treasure trove of information.

But much is still not known about these magnificent natural phenomena. Despite all this attention, hurricanes remain brooding mysteries, hiding their true nature behind their thick, foreboding clouds from scientists seeking to unlock their secrets.

"We actually do not have enough information about the storms when they are out over the ocean," says Da-Lin Zhang, a meteorologist at the University of Maryland, College Park.

Hurricane forecasting is essentially a huge mathematical chess game. About a half-dozen groups of scientists around the world have developed models that predict movements in the weather. These are complex computer programs that crunch millions of numbers and spew out forecasts of all weather, including the tracks of hurricanes.

Zhang and others say that despite huge strides in recent decades, they still do not have all the data these models need to make accurate predictions.

"The aircraft that fly into the hurricane fly in a diamond pattern," says T.N. Krishnamurti, professor of meteorology at Florida State University. "That leaves out a whole lot of places."

Those planes also fly in at 10,000 feet, leaving meteorologists blind to the wind speeds at the ocean surface. The details of the storm's moisture field, which can be affected by the amount of ocean spray the winds are churning up, also remain a mystery.

At Florida State, Krishnamurti and his colleagues compile the forecasts of hurricane tracks from the various models making these predictions and crunch those through yet another computer program designed to take into account each model's biases. The result is a consensus forecast.

But like political polling, all these forecasts, even the consensus one, come with margins of error. That was seen when Hurricane Charley, moving up the west coast of Florida, headed inland well before it was expected. The difficulty, meteorologists say, was that the storm was moving virtually parallel to the coastline so it took only a small change in its direction - well within the forecasts' margin of error - to make a big change in where Charley made landfall.

For Ivan, the models were consistently predicting that it would turn north sooner than it did. The result was the Florida's Keys were evacuated though the storm ended up passing them by on the south side of Cuba.

The forces steering the hurricanes - in these recent cases, a huge area of high pressure over North America - are part of the chess game. If that area weakens, the storm moves north. If it stays strong, the storm continues traveling west. The curve of the Earth also has to be taken into consideration as well as ocean temperature and currents and a huge variety of other atmospheric conditions. All those factors are in constant interplay with the forces in the storm.

The art comes in constructing the models, deciding which weight to give to what factors that are determining the hurricane's path. Once the model is built, it's more of a hard science and the numbers are fed in and the results cranked out.

Sharan Majumdar of the University of Miami says that while hurricane-hunter aircraft have flown into the storms for decades, it is only in recent years that jets have circled the hurricanes at 45,000 feet, dropping dozens of devices that parachute to sea while radioing meteorological measurements back to the aircraft, which send them to forecasters via satellite.

"They are one main reason for the improvement in track forecasts," he says of these flights.

Meteorologists have also been able to get better data from the hurricane-hunter craft by targeting their flight paths at crucial parts of the storm.

But Majumdar contends that the weak link is not the data, it is the models. "I actually think there is a lot of good data from within the storm that is not being used," he says.

The problem, he says, is that the forecasting models are constructed on a global scale and are not detailed enough to give good results on any particular storm in a single part of the globe.

"Models are coming online in about three years, developed at a couple of government labs with a lot of input from university scientists, that can focus on any region around the globe," Majumdar says. "It is basically a high-resolution model."

These models, he says, can crunch all the numbers the satellites and aircraft are gathering. "If we have a very high-resolution grid, it will be better able to use a lot of the aircraft and satellite-borne data from within the hurricane itself."

Majumdar says more effort should go into examining hurricane intensity because very crude meteorological methods are now used to determine and forecast the power of the storms. The focus, he says, should not be just on wind speed, but also on rainfall since flooding once the storm moves inland usually causes more damage than winds do at landfall.

"This new model will be more focused on predicting intensity accurately and predicting rainfall accurately," he says. "There are so many factors in there."

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.