The Future Is Here
We may earn a commission from links on this page

Ancient Shipwrecks Are a Treasure Trove of Climate Data

We may earn a commission from links on this page.

Thousands of sailors perished in stormy seas throughout the Caribbean in the 15th, 16th, and 17th centuries. Surely they would never have guessed that their terrifying ends would help scientists in the future. But creative researchers have found a link between historic Spanish shipwrecks, hurricanes, and the climate.

Their initial research, published last year, takes on added urgency after a devastating hurricane season has sparked conversations about the linkages between powerful storms and climate change.


As a story about scientists and shipwrecks should, it all began over pints of grog: In this case, it was beers following a conference of dendrochronologists, scientists who study tree-rings to mine data about the past. Valerie Trouet, a professor at the Laboratory of Tree-Ring Research at the University of Arizona, Grant Harley, now a professor of geography at the University of Idaho, and Marta Domínguez-Delmás, a dendrochronology researcher and archaeologist at the University of Santiago de Compostela in Spain, were sharing their work on Hotel Congress’ patio in Tucson, Arizona. As they talked, they realized that each had unique pieces knowledge that together, might answer questions about how climate affects hurricanes.

“Grant had talked about how he had collected tree ring data from trees in Big Pine Key, Florida and how [their rings showed] a hurricane signal,” Trouet told Earther. “And Marta described how she dived for shipwrecks. As we talked, we realized there’s an interesting link between these things—shipwrecks and tree rings.”


That link? Both downed ships and ancient trees could tell the scientists something about the history of hurricanes in the Caribbean—going back over 500 years.

This was a pretty big deal, because the official HURDAT hurricane data sets scientists currently work from don’t start until 1850—that’s only 168 years of information. Without a longer period of reliable data, it’s tough for scientists to use that data for what everyone wants—better hurricane predictions. There are proxies that scientists can use to look at hurricane frequency before 1850, but they come from lake and ocean sediments and “they don’t have a data point every year,” said Trouet. “The best you can say is how many hurricanes there were in a century.”

That’s why, when Domínguez-Delmás shared details about the Caribbean shipwreck data with Trouet and Harley, they were so interested. Before modern forecasting, ships by the hundreds were lost to hurricanes, and the worse the storm, the more lives and cargo ended up at the bottom of the Atlantic. So shipwreck data could be directly linked to powerful storms.

Some of these lost ships are really old—the records go back to 1495—and the records are pretty reliable. Investors wanted to know what happened to ships carrying their cargo, so they kept close tabs. Plus, the sailors and other crew aboard lost voyages were someone’s family.


With lots of help from undergrad and graduate students, the scientists went through over 3,000 shipwrecks collected in the book Shipwrecks of the Americas, by Robert Marx. They wound up looking only at Spanish ships, since they were the first country to do the voyage across the Atlantic and therefore had the longest record, as well as some of the most reliable record-keeping.

“We looked only at ships in the Caribbean region, that wrecked during hurricane season and then excluded those that were wrecked for other reasons—war, pirates, etc,” said Trouet. They were left with 657 hurricane-attributable wrecks between 1495 and 1825, with peaks and valleys indicating times when more or fewer ships were lost.


But in order for all this data to be useful for understanding what causes hurricane season to be active or mellow, the scientists needed a continuous data set. The shipwreck details only went through 1825 and the HURDAT set didn’t start until 1851—they didn’t overlap.

That’s where tree-ring data came in. Harley had been taking samples from slash pine trees on Big Pine Key in Florida, some of which were hundreds of years old. He already knew that the growth of these trees, which dendrochronologists can see in the spacing between consecutive rings, isn’t affected by much—not rainfall or temperature fluctuations. However, their growth is suppressed for two reasons: sunlight (the trees shut down growth during mid-November to January each year), and hurricanes.


First, Harley compared the more recent rings from the tree with the record of hurricanes in the more recent HURDAT set. The two datasets matched, showing the tree rings were a good proxy, which he could then use to extrapolate back in time. The three scientists now had a continuous record, from 1495 onward, that told a story about when hurricanes occurred.


What the record showed was that during a period of decreased solar activity from 1645-1715, called the Maunder Minimum, shipwrecks—and hurricanes—were less frequent.

“Climate dynamically it makes sense,” said Trouet. “Your waters need to be a certain temperature in order for hurricanes to happen in the first place—that’s why there is a hurricane season.”


Since tropical storms are born from warmer waters, cooler sea-surface temperatures meant fewer hurricanes. The Maunder Minimum is of special interest to climate scientists because it was part of a larger cold period called the Little Ice Age, which altered weather patterns throughout Europe.

This 500+ years of data can now be added to climate models to help predict what circumstances will cause devastating hurricanes: Trouet, who works with climate modellers, said that the more specific and detailed the data from the past, the more useful the model.


“Now that we know that there were fewer hurricanes during the Maunder Minimum, we can run the models back into the past,” said Trouet. The models that correctly predict the past can then be run forwards to predict what’s coming, and “are more reliable” for the additional detailed information.

While hurricanes are difficult to predict, educated guesses can be made: As global temperatures increase due to climate change, “We can guess that hurricanes will become more intense, but maybe not more frequent,” Harley said. That’s in line with the conclusions of recent National Climate Assessment, which stresses there is still a great deal of uncertainty in how climate change will impact hurricanes. The 2017 hurricane season was a stark reminder of why understanding that linkage is really important.


Harley and Trouet plan to keep plumbing records of the past to shed light on our future climate. They hope to use additional tree-ring data to understand how these historic temperature changes impacted ocean currents—which could have in turn kicked off the Little Ice Age.

One of the wonderful parts about being human is that while we can’t physically travel through time, we can use our ancestors’ lives to inform our own—and maybe predict the future. And it’s heartening to think that maybe all those unlucky sailors whose ships wrecked in Caribbean hurricanes didn’t die for nothing.


As a kid, Starre Vartan argued with her teachers that both science and art were creative endeavors; after 15 years as a science and environment reporter and creative writer, she still thinks that’s true.