A subduction zone caught fragmenting
Researchers have produced detailed images of an oceanic plate breaking into sections beneath the Pacific Northwest, offering evidence that mature subduction zones may shut down gradually rather than stop in a single event. The findings, reported on September 26 and linked to work in Science Advances, concern the Cascadia system off Vancouver Island.
In that region, the Juan de Fuca and Explorer plates move beneath the North American plate. Subduction zones can remain active for millions of years and are associated with major earthquakes, volcanism and long-term reshaping of continents and ocean basins. Geologists have evidence of extinct systems elsewhere, but direct observations of the process by which one winds down have been limited.
The team combined earthquake records with seismic-reflection imaging from the 2021 Cascadia Seismic Imaging Experiment. During that project, sound waves were sent into the ocean floor and their reflections were recorded by underwater instruments carried on a 15-kilometre streamer. Those echoes allowed researchers to map structures below the seabed in a way broadly comparable to medical ultrasound.
The resulting images show large faults and fractures cutting through the descending plate. One mapped feature includes an offset of roughly five kilometres, while a tear extending about 75 kilometres contains both earthquake-producing sections and unusually quiet areas. The researchers interpret the quiet portion as evidence that part of the slab has already separated, because detached rocks no longer remain locked together to accumulate and release stress in the same way.
A piecewise ending over millions of years
The study proposes an episodic process in which individual sections detach at different times, form smaller microplates and establish new boundaries. Transform faults, where parts of Earth’s crust slide past each other, may help cut across the plate. As fragments separate, the remaining slab loses some of the downward force that sustains subduction. Each breakup can take several million years, so the observations do not imply that the entire Cascadia zone is about to fail suddenly.
This model could help geologists interpret fossil microplates and unusual volcanic records elsewhere. Remnants of the ancient Farallon plate off Baja California, for example, have suggested that the larger plate disappeared in stages. An actively fragmenting system near Vancouver Island provides a possible modern analogue for how such remnants formed.
Separation can also create a “slab window,” allowing hotter mantle material to rise through a gap and potentially change magma production. The new mapping therefore contributes to research on plate evolution and volcanism as well as seismic structure. It does not by itself forecast the timing or magnitude of a future Cascadia earthquake; its central result is a clearer physical picture of how one part of a vast tectonic system is slowly coming apart.
