Thinning Ice Caused a Landslide. Then an Avalanche. Then a Tsunami.

This article was originally published by Quanta Magazine.

On September 16, 2023, the world began to rumble. A gargantuan rock-ice avalanche tumbled into the deep waters of a fjord in eastern Greenland, unleashing a mega-tsunami whose initial waves reached a height of 200 meters. The waves scoured the walls of the fjord before flowing into the open sea.

Even for this avalanche-prone corner of Greenland, the collapse and subsequent mega-tsunami were shocking for their speed and ferocity. But what followed was considerably stranger. The collapse triggered a monotonous growl that kept going for nine days straight. It was strong enough to be detected by seismometers all around the world. And yet, when the Danish military visited the scene, they found nothing amiss.

“This is crazy,” says Kristian Svennevig, a geologist at the Geological Survey of Denmark and Greenland and a member of a team that investigated the anomaly. It was like nothing anyone had ever seen.

“Immediately, there was a massive amount of researchers joining forces,” says Finn Løvholt, a tsunami and landslide researcher at the Norwegian Geotechnical Institute, in Oslo. The team quickly grew to include experts from 15 different countries. The Danish navy offered its help. But nothing anyone suggested as an underlying cause—peculiar magmatic movements, jiggling ice sheets—seemed to make sense.

“If we found no other explanation, we would have gone for sea monsters or baby dragons,” jokes Stephen Hicks, a seismologist at University College London and one of the team members.

Read: That wasn’t an earthquake.

Now, as reported last week in the journal Science, a team of 68 scientists has managed to identify the source of the world-wobbling hum: a bizarre natural phenomenon, rhythmically hammering on Earth’s surface like a drum.

“I don’t know if anyone’s seen anything even remotely like that before,” says Ben Fernando, a seismologist at Johns Hopkins University who was not involved with the new study. “It’s a very cool result.”

The joy of discovering something so outlandish has left the team both humbled and overjoyed. “This is probably why a lot of these people went into science,” Svennevig says. “With all our fancy gear and equipment and knowledge, we still, once in a while, find something [we] don’t know.”

On September 16 of last year, at 11:35 a.m. local East Greenland time, a wedge of ice, after years of climate-change-driven warming, thinned away just a bit too much. On the side of a mountain peak atop Dickson Fjord, about 25 million cubic meters of rock—approximately 10 times the size of the Great Pyramid of Giza—thundered downward. As it did, it hit a glacier nestling in a gully, obliterating it.

Lubricated by that ice, and channeled by a narrow gully, this rock-ice avalanche crashed into the fjord at more than 160 kilometers an hour. It “hit the water as a fist,” says Svennevig, a co-author of the new study, and created a mega-tsunami with an average wave height of 110 meters.

Mega-tsunamis usually form when a large, swiftly moving mass plunges into a constricted body of water. The combination of extremely high waves and fast speeds has enormous destructive potential.

As this one careened through a section of the 35-kilometer-long fjord, it stripped away Inuit archaeological sites and 20th-century trapper hunts. It escaped into the sea and partially inundated a scientific and military outpost 72 kilometers away on Ella Island, where its (now far smaller) waves caused $200,000 in damages. Thankfully, nobody was injured or killed.

The initial landslide and the ensuing rock-ice avalanche made a lot of noise. Its acoustic roar was picked up more than 3,300 kilometers away, in Russia, by an infrasound instrument designed to detect clandestine nuclear-weapons tests. Seismometers also picked up a salvo of high-frequency seismic events, along with some low-frequency waves that came from the acceleration and deceleration of the mass grinding on the Earth’s surface.

What came next wasn’t quite so straightforward.

Seismologists picked up a deep hum oscillating at 10.88 millihertz. “This thing was showing up like a really big anomaly, sticking out like a sore thumb,” Hicks says. As they watched it persist for a day, then another, then another, researchers across the world were stunned. “What’s going on here? Why are we still seeing this?” Hicks says, summing up the shock at the time.

By the time the hum decayed, after nine strange days, a variety of experts, including tsunami researchers and deep-Earth geoscientists, had gathered into an expansive chat group on an open-source platform called Mattermost. They debated all sorts of possible explanations, from the routine to the avant-garde.

The seismic signals didn’t match the signature of an earthquake. Perhaps an ice sheet, disturbed by the rockfall, was ringing like a bell. Maybe the landslide had forced a chunk of ice to melt, and that water slipped through a natural piping system in a glacier, turning it into an instrument and creating the geological equivalent of music. Or could the Greenlandic hum be the work of some incognito volcanism?

No evidence to support any of these ideas was forthcoming. People started jokingly wondering if it was aliens, or maybe dragons, having a rave or a tantrum. Some sort of leviathan “always comes up” when abnormalities like this baffle seismologists, Hicks says.

Three days after the great collapse, the Danish navy surveyed the fjord to chronicle the changed landscape. To those scouts, everything seemed tranquil. “It still blows my mind that we’re still seeing this seismic signal going ’round the world, being produced at that time that they sailed there, and they didn’t notice any discernible disturbance,” Hicks says.

Despite a dearth of conclusive evidence, one hypothesis stuck around. Perhaps the avalanche created a standing wave that was sloshing back and forth in the fjord. Such a wave, called a seiche, is common in lakes and harbors, although it typically takes shape in the presence of strong winds.

Then, just three weeks after September’s dramatics, a second (weaker) seismic signal emerged from the fjord and spread through the planet. It also appeared to have been triggered by a landslide, and it was accompanied by a smaller tsunami. Checking through historical seismic records going back to the early 1980s, the team soon found four similar events. If those signals had been from seiches, then perhaps September’s grumble was too.

But there was a glaring problem with the seiche hypothesis. The avalanche and tsunami were large and forceful but ephemeral. It wasn’t clear how they could generate or sustain a nine-day seiche.

In desperation, one member of the team tried to re-create the prolonged sloshing in their own bathtub. Alas, Dickson Fjord isn’t straight and symmetrical, so a bathtub made for a poor stand-in. The noble experiment didn’t work.

Read: The geologists of the future

The team then tried to model the event on a computer. In their first attempts at building numerical simulations of the tsunami, they couldn’t reconstruct the seiche. Then the Danish military gave them access to high-resolution sonar surveys of the fjord’s baroque bottom. That’s when the evidence started to pile up.

Simulations that took into account the varying depth of the fjord showed that the mega-tsunami had stabilized into a seiche with a dominant frequency of 11.45 millihertz—very close to the 10.88-millihertz frequency of the weird hum. The virtual seiche’s nine-day decay matched that of the real thing. The fjord’s wonky geography suggested that the seiche’s energy would only leak out into open waters gradually, helping explain its longevity.

These simulations also revealed the reason the Danish navy failed to see the seiche’s waves in person. The seiche was sloshing the 2.7 kilometers from one side of the fjord to the other every 45 seconds. After just three days, the wave height had decayed to just a handful of centimeters, making it all but imperceptible to anyone present.

The study underscores that our rapidly warming world can affect us in surprising ways. The landslide that kicked off that mega-tsunami was triggered by a melting film of ice. “Climate change is lurking in the background of this story,” Svennevig says. Researchers will now be on the alert for similar warning signs in other glaciated parts of the planet vulnerable to mega-tsunamis.

Nine days of global quivering may be impressive, but it’s not a record. The devastating magnitude-9.1 Sumatra-Andaman earthquake and tsunami in December 2004 shook the planet for 18 days. And it’s thought that the Chicxulub asteroid, which ended the reign of the dinosaurs 66 million years ago, rumbled Earth for months after impact.

The difference with the Greenlandic pandemonium is that, luckily, it caused minimal damage and zero casualties. Instead, Earth presented scientists with a riddle—and they went all in to solve it, simply to satiate their curiosity. “It’s just cool to say: I see a weird signal—what is it?” says Jackie Caplan-Auerbach, a seismologist and geophysicist at Western Washington University who was not involved with the new study. Sometimes, Hicks says, “this kind of science is the most fun.”