The fault off the Pacific Northwest is loaded for a quake that would drop the coastline

Off the coast of Washington, Oregon, and Northern California, a stretch of ocean floor is locked in place and building up strain it has not released in more than three centuries. When it finally slips, geologists expect a rupture powerful enough to drop parts of the coastline by a meter or more in a matter of minutes.

The Cascadia Subduction Zone’s locked interface

The structure responsible is the Cascadia Subduction Zone, a roughly 700-mile boundary where the Juan de Fuca and smaller oceanic plates are sliding beneath the North American plate. Along most subduction zones, that motion happens gradually. At Cascadia, friction has locked the interface between the plates, so instead of sliding smoothly, the two sides are stuck together while the plates keep pushing. That stored strain accumulates over centuries until the interface eventually breaks, capable of producing a rupture near magnitude 9.

What the fault has already done, in 1700

The last time Cascadia ruptured at that scale was January 26, 1700, an event dated with unusual precision not through geology alone but through written records on the other side of the Pacific. A tsunami generated by the rupture crossed the ocean and struck the coast of Japan, where local officials recorded the arrival of an unexplained wave; cross-referencing those records with sediment evidence in the Pacific Northwest allowed scientists to pin the earthquake to that specific night, an account detailed in the record of the 1700 Cascadia earthquake. Coastal geology along Washington and Oregon shows buried forests and marsh layers that dropped suddenly and were flooded by seawater during that event, physical evidence of the same kind of coastal subsidence scientists expect the next rupture to produce.

A recurrence interval measured in centuries, not years

Sediment cores pulled from coastal marshes and offshore turbidite deposits let researchers reconstruct a long history of Cascadia ruptures stretching back thousands of years. That record shows the zone tends to produce a magnitude 9 or larger earthquake on average every 400 to 600 years, though the actual gaps between events have varied widely, from as short as roughly 200 years to as long as 1,000. That variability is part of why the hazard is treated as a matter of probability rather than a fixed countdown: the fault could rupture in a given year, but the average pattern says such an event over any given multi-decade window remains an elevated but non-certain risk. USGS hazard assessments have placed the odds of an approximately magnitude 9 Cascadia earthquake at roughly 10 to 15 percent over the next 50 years.

Why the coastline itself would drop

What makes a Cascadia rupture distinct from many other major earthquakes is the direct, measurable effect on the land itself. As the locked plates suddenly slip, the overriding edge of the North American plate springs seaward and down, a process known as coseismic subsidence. Studies of the 1700 event and the geologic record before it show coastal areas dropping by roughly a meter or more during the rupture, instantly placing what had been dry marshland and forest below the high-tide line. That sudden drop is also what accelerates the arrival of the resulting tsunami: because the seafloor itself is moving, a large wave can begin forming and moving toward shore in a fraction of the time it would take a wave from a more distant earthquake to arrive.

Minutes, not hours, to reach the coast

For communities directly along the outer coast of Washington, Oregon, and Northern California, the proximity of the fault line to the shore means the standard emergency-response assumption for a distant tsunami, that there will be a fifteen-to-thirty-minute warning window, does not apply here. Because the rupture zone runs close to the coastline itself, shaking during a full Cascadia event is expected to last on the order of five to seven minutes, and the resulting tsunami could begin reaching beaches within a similarly short window, leaving little time to move to high ground for people who wait for aftershocks or an official alert. Emergency-management planning documents in Oregon and Washington have used this timeline to justify design standards for tsunami evacuation routes and vertical-evacuation structures rather than relying solely on offshore warning buoys.

Ongoing monitoring rather than a fixed forecast

Because earthquakes of this scale cannot be predicted to a specific day, current research is instead focused on refining hazard maps, shaking models, and building codes based on what a Cascadia rupture would look like if it occurred today. The U.S. Geological Survey’s Earthquake Hazards Program has developed shaking scenarios built from dozens of possible magnitude 9 rupture patterns along the zone, used to stress-test infrastructure, estimate landslide and liquefaction risk, and inform how far inland tsunami evacuation zones need to extend. That modeling work does not shorten or lengthen the odds of an earthquake occurring in any particular year, but it does shape how prepared the region’s bridges, schools, and coastal towns are for the day the locked interface finally gives way.

This article was created with the assistance of AI and reviewed by an editor.

More from Morning Overview