Coupled hurricane-ocean prediction model developed at URI becomes latest tool for National Weather Service

Model expected to increase accuracy of storm intensity predictions by 30 percent compared to the current operational system

NARRAGANSETT, R.I. -- July 18, 2000 -- Listeners tuning in to hear national weather forecasts this hurricane season need to hear this: Trust your meteorologist. The National Weather Service and forecasters now are using even more accurate tools for predicting hurricane paths and intensity, thanks to a model developed by two University of Rhode Island professors that will be added to the national hurricane prediction tool box.

After six years of collaborative research, federal grants of about $2 million, and exhaustive testing through several unpredictable hurricane seasons, the URI Computer Model developed by URI Graduate School of Oceanography Professors Isaac Ginis and Lewis Rothstein has now become an official component of the national hurricane prediction system used to forecast Atlantic and Gulf of Mexico tropical storms and hurricanes.

"With more accuracy in predicting both the track and the intensity of hurricanes, people should pay attention when forecasters report on where a hurricane is going and how strong it is," advises Rothstein. "The margin of error is now 90 miles of coastline when the storm is 24 hours from landfall. Given a full day's lead time, that's very accurate."

This increased accuracy results from the coupling of the URI ocean model with the hurricane model created by National Oceanographic and Atmospheric Administration's (NOAA) Geophysical Fluid Dynamic Laboratory (GFDL). The coupled model now provides a more efficient set of predictors that take into consideration the effects of atmosphere-ocean interaction during storms and more accurate predictions of storm intensity.

The coupled model will be run at NOAA's National Center for Environmental Prediction with the forecasts provided, in real time, to the National Hurricane Center (NHC). The NHC will use this forecast, together with information from other models, to issue an overall forecast.

"Isaac Ginis and Lew Rothstein have done pioneering work in developing their coupled ocean/atmosphere hurricane model," said Max Mayfield, Director of the Tropical Prediction Center/NHC. "The National Hurricane Center is anxious to see the results of their model, and we are hopeful that their results will improve our ability to forecast hurricane intensity."

In 1995 the GFDL system became the new operational hurricane prediction model for the National Weather Service and is currently run at the National Centers for Environmental Prediction (NCEP) for all tropical cyclones in the East Pacific and Atlantic basins during hurricane season. While the current system has performed very well in providing accurate hurricane tracks, it has shown little skill in predicting storm intensity. It seems the difference in intensity may be in the sea-surface temperature.

"Hurricanes are among the most dangerous of our natural hazards," said Ginis. "Each year a devastating toll of destruction is suffered by communities throughout the East Coast and the Gulf of Mexico. Our model helps to fulfill a critical need for forecasters to be able to predict accurately the intensity of hurricanes. This will, hopefully, reduce the damage that occurs after hurricanes make landfall, especially in terms of saving lives."

The GFDL model is presently uncoupled from ocean data, and therefore assumes fixed sea surface temperatures that lead to predictions of more intense storms. Previous studies using coupled hurricane-ocean models as developed by Ginis and Rothstein indicate that the effect of air-sea interaction has a significant impact on storm intensity. This suggests that inclusion of the effect of coupled ocean-atmosphere interactions is critical for accurate intensity prediction of storms.

"While there are other hurricane prediction models, this one is the only one to couple a complex, active ocean general circulation model with a mesoscale atmospheric model," said Rothstein. "It has been chosen as the premier model because it uses the GFDL model as one of its components and significantly improves upon that system with our coupled ocean component."

Rothstein added that the intensity improvements over the GFDL model have been consistent for the past two years with a 30 percent improvement in intensity forecasts. That could translate to one full hurricane category of change that the URI model would forecast as opposed to the GFDL model.

Ginis and Rothstein were recently awarded a $315,000 grant by the National Weather Service to establish a hurricane prediction system over the next three years that will be capable of using both existing and new data from the ocean and atmosphere to enhance hurricane forecasting. In addition, the two have formed Accurate Environmental Forecasting, Inc. to provide forecasting information for the insurance industry. The company recently received funding from the Rhode Island Slater Center for Ocean Technology to further develop and apply this new technology.

After first being tested by the National Weather Service, the science behind the coupled hurricane-ocean model was described in an article entitled "Real-Case Simulations of Hurricane-Ocean Interaction Using A High-Resolution Coupled Model: Effects on Hurricane Intensity" by Morris A. Bender (GFDL/NOAA) and Isaac Ginis (URI), in the April 1999 issue of the American Meteorological Society's journal Monthly Weather Review. (The journal abstract can be viewed online at http://ams.allenpress.com/). Information about the URI model can be viewed at the website: http://tower.gso.uri.edu.

x-x-x

For Information: Jhodi Redlich, 874-2116, jredlich@advance.uri.edu
Lisa Cugini, 874-6642, lcugini@gso.uri.edu