3-D Model to Forecast Hurricane Storm Surge

Media Contacts: Dr. Leonard Pietrafesa, 919/787-6074 (weekends), 919/515-3717 (weekdays) or [email protected]
Dr. Lian Xie, 919/515-1435 or [email protected]
Tim Lucas, News Services, 919/515-3470 or [email protected]

Aug. 30, 1999

Scientists Create 3-D Model to Forecast Hurricane Storm Surge

FOR IMMEDIATE RELEASE

North Carolina State University scientists, working with the National Weather Service office in Raleigh, have developed a three-dimensional forecasting model that accurately predicts coastal flooding caused by a hurricane's storm surge and rainfall.

The model was most recently used to predict where and when flooding would occur in the North Carolina sounds from Hurricane Dennis, which skimmed the coast on Aug. 30. NC State meteorologists/oceanographers Dr. Lian Xie and Dr. Leonard Pietrafesa provided the forecast to the National Weather Service (NWS) on the morning of Aug. 30 and updated it every four hours, to take into account any weakening or strengthening of the storm, or changes in its speed or track. NWS issued coastal flooding warnings based on the NC State model.

"Historically, storm flooding causes the greatest loss of life and property damage when a hurricane strikes," Xie says. "Advance warning of where and when the flooding will likely occur and how high it may go allows emergency management personnel, residents, tourists, marina owners and others to prepare for the worst."

One major advantage of the NC State model is that it takes into account the effects of inland rainfall as well as the surge of the ocean itself. This is something other models, including the one currently used at the National Hurricane Center, can't accurately do.

"Storm flooding can occur well after the storm surge itself recedes. Areas that experienced little flooding from the initial surge can be hard hit when rivers swollen from inland rains begin emptying into the estuaries," says Pietrafesa. In hurricanes Bonnie, Bertha and Fran, for instance, flooding occurred much later in western Albemarle Sound, near the mouths of the Roanoke and Chowan rivers, than in the lower Pamlico Sound. Xie and Pietrafesa's forecasts predicted this and allowed public safety officials to warn communities along the western Albemarle not to let their guard down too soon.

The model is based on a set of computer codes devised by a team of scientists, led by Pietrafesa, in 1986. An early version of the model was first put into use by Pietrafesa for the NWS in 1993 to predict coastal flooding along the Outer Banks from Hurricane Emily. It has since been used to predict flooding from strong winter nor'easters as well.

Forecasts have been accurate to within 10 percent of actual observed water levels. Xie attributes this high level of accuracy to the model's sophisticated computer codes, which allow it to take into account much more detailed data than any other storm flood forecast model. For instance, it computes water-flow speed and direction at 11 different depths, whereas the model developed by the National Oceanic and Atmospheric Administration computes just one average water-flow estimate for the entire water column. "Using our model, we have a much better idea of what's going on near the ocean bottom, which plays a major role in determining how powerful the storm surge will be," Xie says.

The model also features a real-time visualization component that allows users to view real-time computer simulations of where and when flooding will occur on a detailed topographical map of the North Carolina sounds.

Feedback from NWS science officer Kermit Keeter and his colleagues has been instrumental in refining the model, Xie says.

Xie, Pietrafesa and their colleagues have begun work on a new three-dimensional coastal flooding model that will cover not only the North Carolina sounds but the entire coast from the Virginia border to Charleston, S.C. This new model also will be able to incorporate the effects of tides and surface waves, "which can significantly increase the height of the storm surge locally, allowing for better forecasts of flooding on different parts of a coastline or even on different ends of a barrier island," Pietrafesa says.

-- lucas --