Water movement deep inside Earth may explain deep-focus earthquakes

Associate Professor Takayuki Ishii at the Institute for Planetary Materials at Okayama University and an international research team at Peking University and the Center for High Pressure Science and Technology Advanced Research reproduced environments inside subducting plates via high-temperature and high-pressure experiments. The experiments clarified that "water movement" inside plates is deeply involved in the occurrence of deep-focus earthquakes. The findings were published in Nature Communications.

Inside the Earth, when oceanic plates subduct into the mantle, special earthquakes known as deep-focus earthquakes occur at depths of several hundred kilometers. However, because these earthquakes occur at depths where rocks should be difficult to fracture under such high-temperature and high-pressure conditions, their mechanism has long remained a mystery. In addition, some subducting plates bend at depths of around 660 kilometers and exhibit a phenomenon known as plate stagnation, in which they appear to become trapped there, but the cause of this has also remained unclear.

The presence of water within subducting plates is thought to be involved in both phenomena. As oceanic plates descend, they carry water into Earth's interior. Water can accelerate chemical reactions and soften minerals in mantle rocks and has been suggested as a possible explanation for both deep-focus earthquakes and plate stagnation. However, how water moves through Earth's deep interior and influences minerals has not been experimentally verified until now.

The joint research group performed high-temperature and high-pressure experiments reproducing environments inside subducting plates using Kawai-type multi-anvil presses, examining in detail how minerals exchange water with each other.

As a result, in the region at depths of 410 to 660 kilometers, they demonstrated that major minerals inside the plate are dried by having water taken away by special minerals that store water, while on the outside, they are moistened by minerals that released water conversely, meaning that concentration variations of water occur inside the plate. Through this redistribution of water, as high-pressure phase transitions of major minerals advance at the plate center, becoming aggregates of fine minerals causes the plate to soften and deform significantly, and the possibility of generating deep-focus earthquakes was indicated. On the other hand, the outer plate decreases in strength due to water and becomes likely to be deformable, which is considered to play a role in promoting major deformation of the entire plate.

These results explain part of phenomena where plates stagnate or bend at depths. It was also found that in regions deeper than 660 kilometers, such exchanges of water do not occur and that reactions where minerals storing water rapidly dehydrate take place. The deepest deep-focus earthquakes occurring at a depth of approximately 700 kilometers may be triggered by this rapid dehydration reaction, presenting a new mechanism for the origin of the deepest deep-focus earthquakes.

This time, the possibility to explain phenomena occurring in Earth's deep interior, such as deep-focus earthquakes and plate stagnation, from a new perspective of "water movement" was indicated. By having the generation mechanism of deep-focus earthquakes become clearer, it is expected to be useful for modeling earthquake generation and evaluating seismic activity. In addition, with factors influencing how plates subduct and where they stagnate becoming clear, understanding of material circulation inside Earth and evolutionary processes of Earth advances significantly. The findings were published in Nature Communications.

Ishii said, "This achievement is the culmination of five years of research. The idea for how to prove the 'water exchange between minerals' that had eluded direct observation emerged through repeatedly refining the experimental conditions and conducting trial-and-error experiments. When I became convinced that the hypothesis could be demonstrated, my heart started racing".

"Today, AI can generate answers in an instant, but research pursued through persistent trial and error and careful reasoning has a unique appeal. Taking the time to think deeply reveals things that cannot otherwise be seen, and for me, that process is the real excitement of research."

Journal Information
Publication: Nature Communications
Title: Water exchange process and bulk composition regulate slab dynamics and deep earthquakes
DOI: 10.1038/s41467-026-74842-y