AI Identifies 73 Previously Unknown Underwater Volcano Calderas Posing Tsunami Risks

A new artificial intelligence algorithm, originally designed to detect impact craters on Mars, has revealed 73 previously unknown underwater volcanic calderas on the ocean floor. This discovery, led by volcanologist Andrea Vrolino from Paris-Saclay University, triples the number of known submarine calderas worldwide, which were previously fewer than 30. The findings were published recently in the journal Communications Earth & Environment.

The AI scanned ocean floor topography data and initially identified over 87,000 potential structures, which researchers narrowed down to 78 probable calderas, including five already known ones, confirming the algorithm's accuracy. Geographically, eight calderas were found along mid-ocean ridges, nine in volcanic arcs, and the majority, 61, within tectonic plates, often forming underwater mountain chains rather than at plate boundaries.

Underwater calderas can produce highly explosive eruptions capable of generating deadly tsunamis, atmospheric shock waves, ash columns, and massive steam releases. The 2022 eruption of the Hunga Tonga-Hunga Ha'apai volcano in the Pacific Ocean exemplified such risks, causing widespread tsunami damage thousands of kilometers away. The researchers identified seven of the new calderas as high-priority targets for further study due to their location, depth, and shape, which suggest significant hazard potential.

While the study did not determine whether these calderas are currently active, recent research indicates that volcanoes once thought extinct can quietly refill their magma chambers and erupt unexpectedly. The team emphasized that their work establishes a critical foundation for assessing submarine volcanic risks and improving global preparedness, with the goal of expanding this catalog as ocean mapping technologies advance.