Benjamin Horton remembers being in Southeast Asia just months after the devastating 2004 Indian Ocean tsunami. “They were still dealing with a disaster,” he says. “The roads were in a terrible state.” But in those days, the formerly niche field of tsunami research had taken on new urgency. Horton, who studies sea levels at Rutgers University and Nanyang Technological University, was just one of dozens of researchers who came in search of answers: Had this happened before? Would it happen again?

The answers were certainly not to be found in written records or seismometer data. In the short time such data have existed for the Indian Ocean, no one had ever recorded an earthquake capable of sending such a huge wall of water crashing into the coast. The tsunami in 2004 was so deadly because it was so unexpected.

The stratigraphy of the excavated sea cave. The lighter bands are sand deposited by tsunamis over a period of 5,000 years; the darker bands are bat poop. (Courtesy of Earth Observatory of Singapore)

The answer, if scientists were to find it, would probably be in sand. Tsunamis pick up sand from the depths of the ocean floor, depositing it on land as the waters recede. Low-lying coastal plains are good places to look. So are lagoons or mangrove swamps that trap sand. A number of such sites around the Indian Ocean have allowed scientists to begin piecing together a fragmentary history of Indian Ocean tsunamis. To that, Horton and his colleagues now add an exciting new find: a coastal cave in Indonesia containing layers of sand left by tsunamis all the way back to the Stone Age 7,400 years ago.

“It is really a spectacular site,” says Katrin Monecke, a geoscientist at Wellesley College who was not involved in the study, but who has worked on on other tsunami deposits in Southeast Asia. With this cave discovery, scientists have a whole new place they can look for records of past tsunamis.

Horton knew the cave was special the moment he set foot inside in 2011. His colleague, Patrick Daly, an archeologist at Nanyang, had heard about it from locals. The first thing they noticed is that the opening of the cave did not directly face the ocean—a good sign because that positioning slows the movement of water, allowing sand brought by the tsunami to settle in the cave.

Then they stepped in the dark, second chamber. “The next thing you know we were faced with thousands of bats. We were just drenched in bat pee,” says Horton. These bats turned out to be key. Tsunamis had been inundating this cave for thousands of years, during which time bats were also pooping on the cave floor. A tsunami came. Bats pooped. Tsunami, bat poop, tsunami, bat poop, and so on. So when Horton and Daly dug into the sand in the cave, they saw these perfect layers of sand, separated by dark bands of bat poop. “It was a holy grail moment,” says Horton. “We knew we had found something very, very unique.”

Researchers stand in the trench of a sea cave. (Courtesy of Earth Observatory of Singapore)

Over several years, Horton and his colleagues dug six major trenches up to six-and-a-half feet deep. They carbon dated the animal shells and charcoal in the sand layers as well as the bat poop itself. They found, in total, records from at least 11 prehistoric tsunamis, separated by highly irregular intervals. In one case, there was a 2,100 year gap between tsunamis. But within the span of a single century around 1300 BCE, there were four tsunamis. “It shows just how far away we are from being able to predict when an earthquake will hit,” says Horton.

There was one odd hiccup in data. The cave was missing the last 2,900 years of records because, Horton and his colleagues think, the 2004 tsunami actually washed away those layers. The bottom of the sand layer from 2004 was all irregular, as if something (perhaps the 2004 tsunami) had ripped out uneven chunks of the sand bed. This made the team worry that other parts of the tsunami record might be missing. But they did not see the irregular border in any other time period.

For now, the cave is a single data point. The most puzzling and perhaps worrying part, according to Andy Moore, a geologist at Earlham College, is actually the streak of four tsunamis in a single century. “The clustered tsunamis is to me the biggest thing that needs to be clarified because that has real world impacts on how we handle the tsunami hazards on the coast of the Indian Ocean,” he says. How much should countries invest in infrastructure protecting against a potential tsunami? How important is a warning system? “The answer if the tsunami could happen once every few decades to a century, that’s a vastly different answer than [if it happens] every 500 years,” he says.

Horton and his colleagues are continuing to work in the cave. They hope that by looking at the thickness of the sand layers and the grain size of the sand, they might determine how big each of those tsunamis were. They’re also driving down the coastal highway, in search of more caves that can corroborate the records from this one. This highway is relatively new; it was built after the 2004 tsunami.