Scientists have determined that the “swarm” of tens of thousands of earthquakes that occurred near the Greek island of Santorini earlier this year was caused by molten rock flowing through an underground channel over a three-month period.
Using physics and artificial intelligence, researchers pinpointed the cause of the more than 25,000 earthquakes, which propagated horizontally through the Earth’s crust for approximately 20km (12 miles).
The team utilized each tremor as a virtual sensor, subsequently employing artificial intelligence to analyze associated patterns.
Dr. Stephen Hicks from UCL, a lead researcher on the project, stated that combining physics and machine learning in this manner could improve the ability to forecast volcanic eruptions.
Seismic activity began to escalate beneath the Greek islands of Santorini, Amorgos, and Anafi in January 2025, resulting in tens of thousands of earthquakes, many exceeding magnitude 5.0 and felt by residents.
The seismic activity prompted many tourists to leave, and raised concerns among locals about a potential eruption of the nearby underwater volcano, Kolumbo, or a larger earthquake, similar to the devastating magnitude 7.7 quake that affected the region in 1956.
The scientists, who published their findings in the journal *Science*, created a 3D map of the Earth surrounding Santorini. They then mapped the evolving patterns of seismic activity for each tremor, along with the movement and stress within the crust. This process yielded a detailed model of the forces driving the months-long seismic swarm.
The team concluded that the event was driven by the horizontal movement of magma—originating beneath Santorini and the Kolumbo volcano—through a 30km channel situated more than 10km beneath the seafloor.
Researchers estimate that the volume of magma that traversed the crust could have filled 200,000 Olympic-sized swimming pools. These “magma intrusions” fractured layers of rock, triggering thousands of tremors.
Lead author Anthony Lomax, a research geophysicist specializing in seismic activity analysis software, explained: “The tremors act as if we had instruments deep in the Earth, and they’re telling us something.”
“[When we analyse] the pattern those earthquakes make in our 3D model of the Earth, it matches very, very well what we expect for magma moving horizontally.”
For the time being, the researchers believe the event has concluded.
“The magma remained quite deep—more than 8km depth—in the crust,” Dr. Hicks explained. “We know that magma can ascend and erupt at the surface in a matter of hours to days, but because the activity has now died down, we can be almost certain that the melt eventually got stuck and cooled down deep in the crust.”
Volcanoes, however, can enter prolonged phases of unrest and unpredictability that can persist for years, as demonstrated by recent volcanic activity in southwestern Iceland.
These researchers suggest that combining AI with the fundamental physics of crustal movement and stress response could revolutionize the monitoring, understanding, and forecasting of volcanic activity, ultimately enhancing the safety of people in seismically active regions.
“Ultimately, this could be used as a forecasting tool,” explained Dr. Hicks. Whenever a cluster of earthquakes is observed, “that is data that can be used to work out the most likely cause.”
As AI-generated music floods streaming platforms, questions bubble over whether listeners are owed more transparency.
Appearing from the Vatican via video, the pontiff spoke to youth attending a Catholic conference in the US state of Indiana.
The epicentre of the earthquake was about 30 kilometres (18.6 miles) from the capital, Dhaka.
Strong sales at artificial intelligence chip giant Nvidia did little to quell investor worries.
The accountancy giant says artificial intelligence will eventually lead to fewer entry-level positions at the firm.