One morning in September last year, Rudolf Widmer-Schnidrig was doing a spot check of earthquake recording equipment when he saw something extraordinary.
The devices, at Germany's Black Forest Observatory down a repurposed silver mine, detect the most minute rumbles of the planet, allowing seismologists like himself to see signals from earthquakes or other geological events.
But on that day, Dr Widmer-Schnidrig said, "there was a signal which looked different from anything I have seen in 22 years working here at the observatory".
There was a large burst of activity, erratic squiggles on a graph.
And bizarrely, the signal repeated every few hours for nine days.
As it repeated, the signal also tapered off extremely slowly, far slower than expected from something like an earthquake.
The signal could only be caused by something big, but there are very few things we know of that could've produced it.
It might have been a sign of volcanic or tectonic instability or an unknown weapons test. The other exciting possibility was it might have been some rare geological process science had not yet described.
The signal was detected around the world — as far away as Australia and Antarctica — seemingly causing the entire planet to "ring" like a bell.
Dr Widmer-Schnidrig and dozens of seismologists jumped onto online forums to share data and speculate on what happened.
"This is really a very extraordinary occasion to work on such a signal," Dr Widmer-Schnidrig said.
A year on, the scientists unravelled the mystery and reported their findings in the journal Science.
International collaboration
Scientists from 15 countries and 40 institutions pooled their resources and data to figure out what caused the mysterious signal.
Seismologist and study co-author Stephen Hicks of the University College London said with little idea of the cause initially, the researchers referred to the signal as a USO, or "unidentified seismic object".
"It was really like putting a lot of puzzle pieces together," Dr Hicks said.
"I feel like it's similar to those air-crash investigation documentaries.
"One of the first results we had showed the signal was coming from East Greenland."
[map]At the same time the seismologists discovered the signal, the Danish military received reports of a tsunami in the narrow, remote Dickson Fjord, which wiped out a dog sled patrol and research base that was unstaffed at the time.
So they sent a ship to collect evidence of what had happened.
The military found damage to a glacier that could only be caused by a 200-metre-high mega-tsunami in the isolated fjord — one of the highest tsunamis in recent history — 72 kilometres from the island base.
"[The] tsunami was a smoking gun," Dr Hicks said.
"We realised, working with our Danish colleagues, that there's been a huge landslide, so a mountain [called Hvide Støvhorn] had collapsed, rolling down into a glacier.
"That landslide was just able to pick up more momentum, more material from the ice, and then make a giant splash into the water."
Dr Hicks said this caused quite a bit of damage in the fjord, including to some archaeological heritage sites in the area.
"Tourist cruise ships often travel up into these particular fields and thankfully at the time there were no cruise ships in the area, otherwise the results could have been far more devastating," he said.
"So then the mystery was: 'OK, we know a tsunami has happened. But why is this seismic signal still persisting for nine days?'"
Piecing together satellite and drone imagery, before-and-after photos and data, and simulations of the tsunami, the team figured out the signal was caused by a phenomenon called a seiche.
Seiches happen when water is pushed from one side of a body of water, such as a lake, to the other, and rebounds back and forth.
A smaller-scale example is how water sloshes back and forth when you get into a bath tub.
In this instance, the bath was Dickson Fjord, which is 2.7km wide and closed at one end, and the rockslide, the bather.
Through complex mathematical simulations and models, the study researchers calculated there was the initial 200m-high wave, followed by subsequent waves of about 110m, which then stabilised into a seiche with 7m waves.
It was that seiche in the fjord that made seismometers around the world, such as those in Dr Widmer-Schnidrig's subterranean observatory, go haywire.
After further analysis, the team figured out some 25 million cubic metres of rock and ice had fallen into the fjord to create the first powerful wave.
Hrvoje Tkalcic, a geophysicist at the Australian National University who was not part of the study, said it was fascinating to see how seismology and other geophysical sciences were used in the research.
"Not long ago, global seismology was only associated with studying earthquakes," Professor Tkalcic said.
"These days, we see a transition ... to a modern scientific discipline that uses the accumulated knowledge to study the whole Earth as a planet."
But even after the study researchers solved the seiche mystery, questions still remained: what caused the landslide? And would it happen again?
'Ringing' Earth is the sound of climate change
After months of modelling and simulations, plus developing a precise mathematical formula to show how the water sloshed back and forth in Dickson Fjord, the research team turned its attention to what caused the landslide.
Dr Hicks said climate change contributed to the mountain collapse.
"This glacier at the base had been thinning by up to 30 metres over the past couple of decades," he said.
"And so that mountain just wasn't able to be supported anymore."
The study authors pointed out that accelerating climate change would likely cause more massive slides and tsunamis, potentially putting people at risk.
"You could theoretically have similar scale events, particularly as climate change continues," Dr Hicks said.
"[There are] tourist cruise ships going up into these fjords [and] other polar fjords worldwide which are more populated.
"You might only have to have a small instability to really kick off this larger chain of events that evolves into something that can be catastrophic."
Professor Tkalcic thought the world would see similar seiche events in the future for two reasons.
"Glacial melting will increase the rate of ice-related landslides, and due to a unique icy landscape and topography in this part of the world [Greenland], this can lead to tsunamis and seiche phenomena," he said.
"[And] the number of seismic sensors is increasing, which results in more detections of remote natural phenomena we were not able to detect several decades ago.
"Get ready to be surprised by the plethora of behaviour the Earth has in its store."
Dr Hicks also said the September 2023 seiche showed it was important to consider climate impacts beyond the atmosphere or our oceans.
"For the first time this event is showing us that to look at the impacts of climate change, we should also be looking beneath our feet," he said.
A watchful eye
Back at the Black Forest Observatory, Dr Widmer-Schnidrig continues to do his equipment checks, ready to capture any new mystery signals from a changing world.
"If you want to study the Earth's interior, if you want to understand the planet on which we live, you cannot start recording when the [event] has happened," he said.
"When something happens, like this event in Greenland, then you're ready and you have it recorded.
"We should really try the best we can to understand the planet we live on."
