Think Sandy Was Bad? Think Again

As hurricane season begins in the Atlantic, a look to the deeper past
Woods Hole Institute researchers’ study indicates that intensely powerful hurricanes were much more frequent in the region before weather records began to be kept. Jeffrey P. Donnelly

Things could be worse, much worse — and they were on the New England and Mid-Atlantic Coast in the millennia before records about hurricanes were kept, according to an ongoing study by a Woods Hole Oceanographic Institute scientist.

Jeffery P. Donnelly, an associate scientist of geology and geophysics at the Massachusetts institute who visited the South Fork last summer as part of his research, has found that sediment samples from coastal ponds show that huge and powerful storms struck the region with greater regularity and higher intensities during earlier times.

That may seem like heresy in a region that in 2012 witnessed what was dubbed Superstorm Sandy and, further back, endured the catastrophic Hurricane of 1938, but that is what the data indicate. To make matters even more urgent, Dr. Donnelly’s work shows a clear relationship between periods of higher sea surface temperature, like that now linked to global warming, and the most powerful storms.

One Long Island sample provided evidence that at least 12 storms with greater winds and surges than Sandy have struck the region in the past 1,500 years. The most active period he observed was just at the cusp of European colonization, from about 1450 to 1600.

In about 1500 in southeast New England, his team identified five extreme events in four decades — storms of roughly the magnitude of that in 1938. Based on work by other climate scientists, Dr. Donnelly determined that the period correlated well with warmer temperatures in the western North Atlantic. “That’s when we are seeing this big increase in tropical storm activity along the Northeast Coast,” he said.

“The take-home message is that sea surface temperatures matter.”

 Speaking at the Devon Yacht Club in Amagansett last summer to a crowd that included any number of waterfront property owners, Dr. Donnelly described his research in a field he called paleotempestology, or the study of storms from before recorded history.

He said he had conducted a sampling visit in July to Fort Pond in Montauk, where he and his team deployed the smallest and most low-tech of Woods Hole’s fleet of research vessels. Usingtwo canoes joined together by a simple plywood deck, the team drove a long metal tube into the pond’s bottom, pulling up a column of sediment whose layers, they hoped, had a story to tell. This was a process they had repeated up and down the New England and northern Mid-Atlantic coast.

The work was motivated in part by representatives of the re-insurance industry who got in touch with Dr. Donnelly after 1992’s Hurricane Andrew caused $26 billion in damage, mostly in Florida.

These large insurers, who underwrite the activities of the smaller companies that provide individual policies, were uncomfortable gambling huge sums of money based on only about 160 years of historical weather records. Was there some way to get a longer view, they wondered. And, perhaps more important, what might the changing climate mean to their bottom line?

Dr. Donnelly’s study is relatively easy to understand. He and his team members take samples from the muck at the bottom of coastal ponds. Back at Woods Hole the layers can be analyzed and carbon-dated with precision.

Ordinary years are marked by the steady accumulation of muddy sediments. But then occasionally these are interrupted by irregular layers of sand, evidence of powerful events that carried material from the shoreline up and over barrier dunes. Together with the date evidence, the sand lenses provide an unarguable record of the most devastating storms.

By taking samples from ponds up and down the coast, Dr. Donnelly’s team is able to estimate a storm’s intensity by comparing the size of sand grains, and get a good idea of where its center struck land.

Dr. Donnelly said the longest coastal record his lab had analyzed went back 5,500 years, from a site in Puerto Rico. A student of his working in the Bahamas took a sample going back more than 7,000 years. By that measure, those in the Northeast, which span about 2,000 years, are relatively modest, but they still provide a wealth of data.

The information gleaned from the pond cores can be compared to storms known from meteorological records, as well as to results from other long-term studies, for clues to their connection to climate and global weather patterns. Other work shows close correlation with El Nino cycles and other complex factors.

The implications for policy-makers concerned about manmade climate change is clear, he said. Storms like Sandy are hardly as bad as extreme weather events can get.

An 1821 hurricane he studied left sand deposits in 14 coastal ponds between Long Island and southern New Jersey; during Sandy, telltale sand ended up in only one of those locations. Adding to the sense of alarm, sea level has risen about 12 to 16 inches since the early 19th century, allowing hurricanes to ride into the coast that much higher, and with greater impacts.

Last year’s Fort Pond sample, as it turned out, was not all that valuable relatively. Within the last couple of thousand years, Dr. Donnelly said, the northern portion of the pond was open to the waters of Block Island Sound, obscuring the geologic record.

Predictions for the 2014 hurricane season are for a below-average number of storms.