Scientists have discovered that a “swarm” of tens of thousands of earthquakes near the Greek island of Santorini earlier this year was caused by pumping molten rock through an underground channel over three months.
They used physics and artificial intelligence to figure out exactly what caused more than 25,000 earthquakes that traveled about 20 km (12 miles) horizontally through the Earth's crust.
They used each of the tremors as virtual sensors and then used artificial intelligence to analyze the patterns associated with them.
One of the lead researchers, Dr Stephen Hicks from University College London, said this combination of physics and machine learning could help predict volcanic eruptions.
Seismic activity began to intensify beneath the Greek islands of Santorini, Amorgos and Anafi in January 2025. Tens of thousands of earthquakes have occurred on the islands, many of which were greater than magnitude 5.0 and could be felt.
Many tourists fled and locals feared that the nearby undersea volcano Columbo might be about to erupt, or that it was a seismic prelude to a larger earthquake similar to the devastating magnitude 7.7 quake that struck the same region in 1956.
Scientists, publishing their findings in the journal Science, have created a 3D map of the Earth around Santorini. They then mapped the evolving patterns of seismic activity from each shock, as well as the movement and stress in the Earth's crust. The result was a detailed model of what exactly caused this months-long seismic swarm.
The team found that the event was caused by the horizontal movement of magma – from beneath Santorini and Volcano Columbo – through a 30-kilometer-long conduit that lies more than 10 km below the seafloor.
Researchers estimate that the volume of magma that passed through the Earth's crust could fill 200,000 Olympic-sized swimming pools. These “magma intrusions,” as they are called, burst through layers of rock, causing thousands of tremors.
Lead study author Anthony Lomax, a research geophysicist who develops scientific software for analyzing seismic activity, explained: “The sounds act as if we have instruments deep underground and they are telling us something.
“[When we analyse] The pattern that these earthquakes create in our 3D model of the Earth matches very, very well what we would expect from the horizontal movement of magma.”
For now, researchers say, it looks like it's all over.
“The magma remained quite deep – more than 8 km – in the Earth's crust,” Dr Hicks explained. “We know that magma can rise and erupt to the surface within hours or days, but since the activity has now died down, we can be almost certain that the melt eventually became trapped and cooled deep in the Earth's crust.”
However, volcanic eruptions can enter prolonged phases of unrest and unpredictability that can last for many years. This has been demonstrated by recent volcanic activity in southwest Iceland.
And these researchers say that using AI, combined with the fundamental physics of how the Earth's crust moves and responds to stress, could transform the ability to track, understand and even predict volcanic activity. This could help keep people safe in seismically active parts of the world.
“This could eventually be used as a predictive tool,” Dr. Hicks explained. Whenever we see a cluster of earthquakes, “that's data that can be used to determine the most likely cause.”
