CONWAY, S.C. -- As hurricane season approaches its peak, researchers at Coastal Carolina University on Thursday unveiled a new hurricane model program named after a massive Category 4 storm that smashed into South Carolina a quarter-century ago.
Scientists say the model called Hurricane Genesis and Outlook, or HUGO, provides predictions of where hurricanes will hit coupled with a better estimate of storm surge and flooding. Hurricane Hugo hit South Carolina with 135 mph winds in 1989.
Len Pietrafesa, the head of the research team, said the model predicts both the number of storms expected in any given season and also the probability of a storm making landfall along the East Coast or Gulf Coast. For this season, the model predicts the highest probability is that one storm will make landfall; the second-highest probability is two storms.
In addition, he said, the model predicts track and intensity of hurricanes five days from landfall, reducing the so-called cone of uncertainty of where the storm will hit and providing better estimates of storm surge and flooding. The model considers statistical data from storms dating to 1950.
"These are outlooks and flooding projections, I might say, that other agencies and institutions have been doing for many years. So what is new with what CCU has put onto its computational platform?" Pietrafesa asked.
"Well, we have included a new prediction in outlook which is a landfall prediction and we have included a new flooding, surge and inundation model," he said. The model, he said, includes 22 complex factors including such things as oceanic and atmospheric activity as well as shoreline development.
Pietrafesa is a former chairman of the National Hurricane Center External Advisory Board and hopes the model will one day be used by the center in its forecasting. That's something that could take years.
There are numerous hurricane models, and the Hurricane Center's chief of forecast operations, James Franklin, who has not seen the Coastal Carolina model, said they are tested extensively.
"If they are interested in making their model for us to use, first we would ask for a large set of reruns, most likely three years," to see if the model accurately predicts past storms, he said. "If that looks promising, we probably would be interested in seeing some live trials and have them run it for a year live. We wouldn't put it in front of the forecasters until we had seen the results. We just won't just throw things in front of the forecasters."
In the meantime, the model's predictions are available to emergency officials and whoever else may want to use them on the university's website.
The Hurricane Center relies on seven main models and reviews various research models each year, Franklin said, adding there will never be a perfect model.
"We can't measure temperature in the atmosphere everywhere. We can't measure wind in the atmosphere everywhere. We're never going to know with perfection what the current state of the atmosphere is right now," he said. "If we can't know what the state of the atmosphere is right now, you can't possibly have a perfect forecast. Your starting point will always be flawed, and those flaws will grow with time."