Duke Geologist Decries Using Engineering Models To Predict Natural Phenomena

"Engineering models of Earth surface phenomena do not work, and I think it is very important that society recognizes this," said Orrin Pilkey, a professor of Earth and ocean science at Duke's Nicholas School of the Environment. "I think mathematical modeling in this sense is damaging our society."

Pilkey spoke in an interview before his talk on the same subject Monday at the annual meeting of the American Association for the Advancement of Science.

He is the outspoken director of Duke's Program for the Study of Developed Shorelines and the co-author of The Corps and the Shore (1996, Island Press), a book that criticizes the U.S. Army Corps of Engineers' use of predictive mathematical models such as "GENESIS" and "SBEACH"to model beaches.

Such models use mathematical equations to represent such factors as wave characteristics, the density and porosity of beach sand, the density of sea water and the beach's slope. The intended purpose is to predict how natural or artificial beaches will respond to various kinds of coastal engineering.

Some engineering models -- such as those assessing the performance of construction materials in a bridge or building -- are useful, Pilkey acknowledged. "Modeling steel and concrete is fine," he said. "But now engineers have stepped into natural systems and they want to model beaches. Well, they can't do that.

"In general, they can't model any natural systems with precision. But because beaches are so dynamic and because they can change so suddenly, it is very obvious and easy to see why modeling fails for them."

Pilkey focused on expensive "nourishment" projects where sand from another location is pumped or trucked in to embellish an eroding beach. Corps engineers use mathematical models to assure beach communities that these projects will have a definite "lifetime." But reality is quite different from these models, he argued.

"What happens to a beach depends on randomly occurring storms, and every storm is different," he said. "You have some big storms that don't do anything. Then the next big storm will do a tremendous amount of damage."

Beaches can behave chaotically, which -- according to the emerging guidelines of chaos theory -- "have a huge dependence upon initial conditions," added Pilkey, who said his "favorite initial condition" is a layer of shells that can cover some beaches.

"If you have a shell layer, the beach won't move until you are well into a storm and the energy gets really high," he said. "Whereas, with a beach that is less shelly, its sand will move with the waves from the beginning of a storm. So you have an entirely different response to a storm just because of a layer of shells."

Pilkey said such models also do not keep up with the pace of scientific discovery.

For instance, he said, coastal researchers learned only comparatively recently that "rocks underlying the shore face control the erosion rates of beaches. These models also don't take into account the strong bottom currents that we didn't know much about 20 years ago."

Pilkey stressed that he does not object to using "basic" scientific models to study nature, because those address the questions "how" or "why." By contrast, the engineering models he opposes "ask the questions "where," "how much" and "when," he added.

"For example, "When will that beach disappear that we just nourished?" he said. "Or, "How much sand will be needed to make it last for X numbers of years?"; and "Where is the shoreline going to be 25, 50 or 100 years from now?"

Pilkey proposes that what is called beach erosion is actually tied to a natural process linked to rising sea levels. He has attracted controversy by arguing that coastal engineering can exacerbate erosion and that some measures -- especially seawall construction -- can actually destroy a beach.

While he acknowledges that beach nourishment projects can work for a while, he contends that their predicted lifetimes are often wildly optimistic, forcing beachfront communities to find the money for yet more replacement sand.