What Are The Chances Of The U.S. Being Hit By A Catastrophic Hurricane? LSU Professor Takes A New Field Of Hurricane Research By Storm

Liu's latest study provides concrete information about the probability of catastrophic hurricanes making landfall in the U.S. and shows that the hurricane activity of the most recent millennium has been mild, both in frequency and intensity of storms.

The chances of a catastrophic hurricane hitting the U.S. Gulf Coast are about once every 300-600 years, the study shows. The findings provide a specific number range for the first time, answering what Liu calls the "$30 billion question" asked by insurance companies and home owners about the probability of U.S. hurricane landfalls. Liu said Hurricane Andrew, a category 4 storm, caused $30 billion in damage.

Liu presented his findings Feb. 19 at a hurricane symposium that is part of the annual meeting of the American Association for the Advancement of Science. The 2000 meeting was held Feb. 17-22 in Washington, D.C.

Liu, LSU's James J. Parsons Professor of Geography, is one of nine leading hurricane experts who spoke during the day-long symposium titled "Frontiers in Hurricane Climate Research." He delivered his presentation, "Paleotempestology: Reconstruction of Past Hurricane Landfalls from Sedimentary Proxy Records," during the morning session and will serve as chair of the afternoon session. Liu is also an organizer of the symposium.

By examining sediment in coastal lakes and marshes along the coasts of the Gulf of Mexico and the Atlantic Ocean, Liu has been able to reconstruct the frequency and intensity of hurricane landfalls dating back some 5,000 years. This cutting-edge field of research is known as paleotempestology, or the science of studying storm activities of the past. It is a relatively new field of study, and Liu is taking it by storm.

"Our records of hurricanes only go back about 150 years, and that is a short time to observe these storms and to predict the probability of being hit by a catastrophic hurricane," Liu said. "Hurricane Camille in 1969 is the only category 5 hurricane that hit the U.S. mainland during that time. We need to look back longer to put the present climate into perspective. If we don't look far enough back in the past, we don't know what could be in store for us in the future."

The storm surge that accompanies a hurricane washes up onto the shoreline when a storm makes landfall. This "storm overwash" carries sand from the beach and dunes into coastal lakes and marshes, forming a sand layer. With each new hurricane comes a new layer of sand. By digging up the sediment and using radio-carbon dating to date the layers, Liu was able to determine the number of hurricanes that have hit the coast for the past five millennia.

Liu has been performing this procedure on the coast of the Gulf of Mexico for the past 10 years, examining more than 16 sites from Florida to Texas. He has been working on the Atlantic Coast for the past two years, where he studies six sites from Virginia to Massachusetts.

Although Liu's findings show that catastrophic hurricanes -- those of categories 4 and 5 -- make landfall in the U.S. about once every 300-600 years, he said the numbers are higher along the Gulf Coast than the Atlantic Coast. His findings also show that there are fluctuations in hurricane activity from one millennium to another, just as there are from one decade to another. These fluctuations are the result of varying climatic patterns that cause hurricanes to be mild and infrequent during some periods in history, and to be catastrophic during other time spans. While the 300-600-year time frame might be considered good news to some coastal residents, the bad news is that hurricane activity during the past millennium, when compared with other periods of time, has been mild and inactive, Liu said.

"People think Camille and Andrew were devastating, but we haven't seen anything yet," Liu said. "If we switch back to a more active state, the U.S. could be hit a lot more frequently than we've seen in our lifetimes. There is a very distinct millennial-scale variability, and in the past 1,000 years, there has been a very low incidence of major hurricane landfalls along the Gulf Coast."

Liu said such information is vital to coastal-development planners and insurance-risk assessors, not to mention the millions of people in America who live in coastal zones.

"Insurance companies charge us based on a probability that they came up with," Liu said of people living along the U.S. coastline. "Until now, there was no empirical basis for estimating the probability for catastrophic hurricanes. Now we have numbers that are scientifically validated."

Liu said his study was funded by both the scientific community and the insurance industry, exemplifying an unusual union between science and society. Both the Risk Prediction Initiative, housed at the Bermuda Biological Station for Research, and the National Science Foundation have funded his work.

The symposium dealing with hurricanes was included at the AAAS meeting because of the increase in hurricane-induced losses of property during the past decade, Liu said. He pointed out that the 1970s and 1980s were periods with a relatively low incidence of intense hurricanes. During that time, an influx of people to coastal areas caused huge population and property increases along the U.S. coastline. When hurricane activity increased in the 1990s, there were more lives and property to be lost than ever before. Liu said these "inter-decadal" variations in hurricane activity can give people a false sense of security.

"When we go a long time without a hurricane, we forget the lessons we learned from storms like Camille," Liu said.

He said the symposium will examine the new frontiers of hurricane research, such as paleotempestology, as well as the climatology of hurricanes and planning and evacuation issues. The speakers will address past and potential future changes in hurricane landfalls, the relationship between hurricane activity and the condition of the oceans and the atmosphere, the ability of meteorologists to predict hurricane landfalls and how the public responds to the risk of a hurricane.

Other presenters at the symposium included researchers and scientists from the National Hurricane Center, the Massachusetts Institute of Technology, the National Oceanic and Atmospheric Administration, the Bermuda Biological Station for Research, Florida State University, the University of Rhode Island and the risk-management industry.

The AAAS annual meeting provides an opportunity for scientists from around the U.S. to network, discuss research topics and findings and announce major scientific discoveries.