Predicting Storm Surges

Forecasters know Hurricane Rita’s winds will generate a tremendous storm surge, but just how high and how wide the devastation will be remains hard to predict as the powerful storm hooks toward the Texas-Louisiana border.

The dome of water pushed toward shore by the winds is the greatest potential killer for any hurricane, experts say. Emergency managers know that the only way to be safe from storm surge is to get out of the way _ going higher in a building that may get pounded apart by waves offers no guarantee of survival.

Deaths from the sea during a hurricane became rare since Hurricane Camille scoured the Mississippi coast in 1969, partly because few catastrophic storms had hit populated areas but also because models detailing the threat had gotten better.

Then came Hurricane Katrina. Most of the more than 1,000 confirmed storm-related fatalities in Mississippi and Louisiana resulted from direct ocean flooding or from secondary flooding started by storm surge battering the levees around New Orleans.

Katrina’s reach was quite wide and extended over relatively shallow waters most susceptible to storm surge, sending water up to 30 feet above sea level in some areas and 10 or more miles inland. While the surge was worse in places than forecast, the bigger problem in Mississippi appears to have been people either not respecting the forecasts or believing that if a structure had weathered Camille, it would be fine during Katrina.

In fact, while Camille was more powerful than Katrina, the Aug. 29 storm had a much bigger wind field, and “we’ve learned the size of a storm out over the ocean matters almost as much as the strength of the winds when it bumps into land,” said Ed Rappaport, deputy director of the National Hurricane Center in Miami.

The center uses a computer model called Sea, Lake and Overland Surges from Hurricanes (SLOSH) to calculate the possible maximum surge from a storm. It takes into account the depth of the water just offshore, as well as the height of barriers like dunes and seawalls, to project how wind-driven water will move.

Emergency managers in most communities take the SLOSH models and further refine the predictions using more detailed maps to determine just which areas need to evacuate.

“Any forecast we do is dependent on the track, the intensity and the size of the storm,” said Will Shaffer, chief of the forecast evaluation branch of the National Weather Service. “When we get all those elements right, the model gives us a pretty good estimate of what the surge is.”

Shaffer said the surge forecasts are generally accurate to within 20 percent, according to post-storm surveys that use tide gauges and high-water marks on surviving structures.

“It’s hard to do much better than that, except to the extent we improve our ability to forecast the track and intensity,” Shaffer said.

In fact, forecasters have improved their margin of error for predicting, 24 to 48 hours in advance, where a hurricane will make U.S. landfall down to about 100 miles. The geographical specificity is a 50 percent improvement over what was possible a decade or two earlier.

Forecasts of intensity have improved only by about 10 percent in the same period.

Even so, a change in a hurricane’s track of even tens of miles can change the effect of storm surge pretty dramatically, Shaffer said. For instance, as Hurricane Ivan bore down on the central Gulf Coast last fall, the forecast track 12 hours before landfall took it just west of Mobile Bay, Ala., and anticipated a surge of 12 to 14 feet in the bay, while Pensacola, Fla., was expected to get a 3-5-foot surge.

But Ivan landed a bit farther east and packing a larger wind field than expected, with the result that Mobile Bay had a storm surge of just 3 or 4 feet, and Pensacola’s Escambia Bay was hit with a 12-foot surge.

On the Net:

(Contact Lee Bowman at BowmanL(at)