New Colorado State Landfall Probability Prediction Model Gives Likelihood of Hurricane Striking U.S. Coast at 66 Percent in 1998

There is a 66 percent chance that an intense hurricane will strike somewhere along the U.S. Gulf and Atlantic coastline in 1998, according to predictions by a new model of hurricane landfall probability developed at Colorado State University.

The statistical model, developed by William M. Gray, professor of atmospheric sciences at Colorado State, will be released Friday. It divides the Gulf and Atlantic coastlines from Brownsville, Texas to Eastport, Maine into 11 major segments and predicts landfall probabilities. The coast is further divided into 100 kilometer (62 mile) sections and a landfall probability is assigned to each.

For example, Gray said, Miami faces a 3.3 percent probability that an intense hurricane (with winds of 111 m.p.h. or higher) will come ashore this year in the 100-kilometer segment in which it lies. Probabilities for other major cities are: New Orleans, 2.4 percent; Houston, 2.5 percent; Charleston, 1.6 percent; and New York City, 0.9 percent.

Gray’s Colorado State research team also issued its final 1998 hurricane season update today, holding its forecast to 10 tropical storms, six hurricanes and two intense hurricanes (see accompanying report and press release). Gray and his colleagues have been predicting numbers of tropical storms, hurricanes and intense storms with above-average accuracy for the past 14 years.

Gray’s new analysis reveals that some areas of the U.S. coast never have experienced landfall by an intense hurricane (one with sustained winds of 111 m.p.h. or above) this century, while other regions have been struck repeatedly. Since 1900, 14 intense storms have hit the Florida peninsula from south of Naples through the Keys and up the east coast to Fort Pierce. This century, 11 intense storms have come ashore from the Georgia-South Carolina state line north to Cape Hatteras, 18 along the Lafayette, La.-to-Panama City Beach coast and six between New York City and Cape Cod.

However, not one intense hurricane has struck between Fort Pierce and the Georgia-South Carolina border this century, nor has one made landfall between Cape Hatteras to just west of New York City.

"The reason you don’t have storms landfalling there is because there are no trade winds coming in from the east at that point, and storms typically ‘recurve’ and parallel the coast in these areas," Gray said. "The steering currents tend to be more from the south in that area.

"But south Florida gets the trade winds blowing in from the east."

Prevailing winds and other climate factors also help protect a coastal span that straddles the Texas-Louisiana border and areas extending east and southeast from near Tallahassee on Florida’s Gulf Coast and from Cape Cod to the Canadian border, none of which has experienced an intense landfalling hurricane this century.

Lesser storms – named storms and stronger hurricanes rated 1 or 2 on the 1-5 Saffir-Simpson system – are more widely distributed.

"If you include lesser storms, you do fill in the map of the coast," Gray said. "And tropical storms come in all over the place."

"I don’t think a lot of observers, even insurance companies, realize that hurricane-spawned potential areas of damage vary widely up and down the coast."

There are two key indicators in the landfall probability model (see accompanying press release): a statistical average of four to six measures of overall Atlantic Basin hurricane activity for a given year called "Net Tropical Cyclone" (NTC) activity; and the "Atlantic conveyor belt," or thermohaline circulation. This system moves Atlantic Ocean water northward from the vicinity of the Caribbean to an area east of Greenland. There, the current sinks to deep levels, moves southward and flows into the South Atlantic Ocean and beyond.

The conveyor belt phenomenon is measured by long-term change in North Atlantic sea surface temperatures and salinity. Cooler temperatures and lower salinity, which prevailed from 1900-1925 and 1970-1994, inhibit hurricane formation and landfall. Warmer temperatures, prevalent from the late 1920s to the late 1960s, and more salinity produce a more intense season. Since 1994, North Atlantic sea surface temperatures have been warming and ocean surface salinity is higher.

"We are pretty confident that certain climatic factors have changed and that we’re going to see more landfalling hurricanes," Gray said. "We think the Atlantic will be moving into a more active era for the next 20 years or so."

"When the climate signals are favorable for hurricanes you get a lot more landfall," Gray said. "And once the climate signals are really good, the number of landfalls goes up disproportionately. Big storms have the best climate signals."

Category 3 or higher storms constitute only about a quarter of the tropical cyclones that have made landfall this century (there have been 72), most of them causing from 10 to 100 times as much damage as a typical category 1 hurricane when they reach shore. An estimated 85-90 percent of all tropical cyclone damage occurring along the southeast U.S. coast has been caused by category 3-5 storms making landfall.

Gray said the new model’s predictions will be incorporated into the Colorado State research team’s annual pre-season hurricane forecast, issued each December and updated in early April, June and August, but that the data offers only very general information.

"Coastal residents should be aware of these changes in the climate signal that can affect their odds of being hit," he said. "But this is a statistical model, and these predictions are general. The major damaging winds don’t cover a very big area, no more than 100 miles or so of coastline. Really intense winds cover even less, say 50 miles.

"Even in an active year, the probability of being hit by an intense storm is really low."