According to a new study i earthquakes that occurred in 2025 in Santoriniin Greece, would not be of tectonic origin, i.e. due to movement along faults associated with the margin between two lithospheric plates, as had been hypothesized. Instead, it would have been the cause of them an infiltration of magma in the rocks, which advanced in waves without however reaching the surface. The discovery, published in the journal Scienceis the result of a collaboration between ALomax Scientific (France), the Aristotle University of Thessaloniki (Greece) and University College London (UK). The researchers came to this conclusion analyzing over 25,000 earthquakes occurred in Santorini in 2025, with the help ofartificial intelligence.
Why Santorini was shaking: the magmatic origin of the 2025 earthquakes
The research team analyzed data recorded by seismometers between the end of January and the beginning of March with the machine learning (or machine learning), a branch of artificial intelligence. It was thus possible to obtain a database of tens of thousands of earthquakes, hundreds of which with magnitudes greater than 4.5. Analyzing the spatial distributionresearchers were able to determine the cause of the seismic activity. In particular, as study co-author Stephen Hicks explains, earthquakes were considered as “virtual stress meters, i.e. clues to how stress was changing undergroundHicks adds: “This gave us a robust, high-resolution picture of what was happening, allowing us to rule out fault slip as the main cause of the earthquakes”.
Seismic activity was found to be extensive up to 50 km north-east of Santorini And up to over 15 km deep. The distribution of earthquakes highlighted the presence of a dike, an intrusion of magma inside the rocks. The intrusion developed from an underground magma reservoir that connects the Santorini volcano with the nearby Kolumbo underwater volcano. The researchers managed to trace the way in which the magma advanced underground, propagating horizontally: the flow occurred in wavesin the form of impulses, opening fractures and stopping after the rock breaks, and then starting to exert pressure and advance again.
It was these pulses that triggered earthquakes in the surrounding crust. Fortunately, the pressure and density of the magma were not such as to allow it to reach the surface and cause a volcanic eruption. The changes in the Earth’s surface recorded by satellite, however, have made it possible to estimate that the volume of the magma of the intrusion could have filled 200,000 Olympic swimming pools. The goal for the future is to build a 3D image of the crust during the months of intense seismic activity in 2025 in Santorini.
Future prospects in monitoring Santorini
In the case of Santorini, machine learning made it possible to identify a particular mechanism by which magma can be transported, which could also characterize other volcanoes in the world. Furthermore, he highlighted the volcanic danger of the area and consequently the need for continuous high resolution monitoring. With current technologies, we are able to measure vibrations and deformations on the Earth’s surface. Machine learning instead made it possible to obtain a high resolution picture of the situation even in depth. The technique used in this study, explains Stephen Hicks, “it could be applied to future earthquake swarms in near real time and could allow us to better predict the likelihood of volcanic eruptions or stronger earthquakes. If we applied our technique to similar earthquake swarms in the future, we could pinpoint precisely where magma might escape and potentially in how much.”.
