The U.S. Geological Survey puts the chance of a magnitude 9 earthquake rupturing the full length of the Cascadia megathrust at about 15 percent over 50 years, using a model that counts the time elapsed since the last great quake on January 26, 1700. A simpler model with no memory of that date gives 10 percent for the same window, and both rest on a recurrence interval of about 500 years for a full-margin rupture.
Help yourself and the planet: The Cool Down makes it easy to help yourself while helping the planet, with smart, simple changes to save more and waste less in your everyday life.
We may earn a commission if you sign up.
The percentages are absolute probabilities, not relative ones, and they describe one specific kind of event: a rupture spanning most of the coastline, along the Pacific Northwest coast. Partial ruptures, which break only part of the margin, are counted separately, and the gap between the two kinds of event explains most of how the numbers are built.
Two models, 10 percent and 15 percent
The figures come from USGS Fact Sheet 2025-3050, “Earthquake probabilities and hazards in the U.S. Pacific Northwest,” whose suggested citation names Wirth, Frankel, Sherrod, Grant, Dunham, Stone and Grossman as authors. It covers California, Oregon, Washington and southern British Columbia. The simpler calculation is a Poisson model, P = 1 − e^(−t/τ), where t is the 50-year window and τ is the mean recurrence time. With a 500-year recurrence that works out to roughly 10 percent. The fact sheet states the model’s central assumption plainly: “Assuming a time-independent model implies that an earthquake is equally likely to happen on any given day.”
The authors call that assumption likely incorrect for any individual fault but reasonable when dates are uncertain or when several faults are combined, which is why it stays in the analysis as a baseline.
For full-margin Cascadia ruptures the authors also apply a time-dependent model, a Brownian passage-time calculation from Matthews and others (2002), that accounts for the 1700 earthquake. More than three centuries have passed since that date, and the model treats the elapsed time as raising the odds of the next rupture within the 50-year window. The result is the figure of about 15 percent. In plain terms, the same fault gets roughly a one-in-seven chance under one model and one-in-ten under the other, and the fact sheet publishes both rather than choosing a single answer.
The 500-year recurrence interval that feeds both models is itself an inference. The National Seismic Hazard Model bases it on evidence from coastal subsidence, tsunami deposits and turbidites, the underwater sediment layers that great earthquakes leave behind.
The 1700 rupture and the older record
The fact sheet says the last great Cascadia earthquake struck on January 26, 1700, with an estimated magnitude between 8.7 and 9.2. Several independent records fix the date: tree-ring dating shows that coastal trees died in the winter of 1699 to 1700, land-level change and seawater inundation are preserved in the ground, and a written record of a tsunami in Japan gives the day itself. A 2003 analysis by Satake, Wang and Atwater of Japanese tsunami descriptions found a most likely magnitude of 9.0, with about 19 meters of slip on an offshore zone 1,100 kilometers long, and reported waves of 1 to 5 meters at seven Japanese shoreline sites.
Earlier ruptures were not all alike. In a 2006 study in Quaternary Research, Nelson, Kelsey and Witter identified about 16 events in the geologic record, eight long ruptures including 1700 and eight more limited ones, and concluded that Cascadia produced great earthquakes of variable magnitude. The northern plate boundary, they wrote, usually breaks in long ruptures during the greatest earthquakes, while the southernmost segment varies more in both length and recurrence. That variability is the reason the hazard model keeps full-margin and partial-margin ruptures in separate columns.
Southern Cascadia and the 30 percent figure
The fact sheet gives a different figure for the south: a time-dependent model puts the chance of a magnitude 8 or greater earthquake in southern Cascadia at about 30 percent in 50 years, counting both full and partial ruptures. Partial ruptures there may recur every 800 to 1,000 years, and combined with full-margin events the overall recurrence in southern Cascadia is about 300 to 350 years. The shorter interval is why the southern number runs higher than the coast-wide magnitude 9 figure.
The authors are open about the limits. They note that it is unclear how to apply a time-dependent model in a region that may produce a range of magnitudes, and that evidence for partial ruptures in northern Cascadia exists but its recurrence is hard to pin down. Even if full-margin ruptures turn out to be series of partial ones, they write, individual sites along the coast would still experience a great earthquake of magnitude 8 or larger about every 500 years.
Land subsidence and tsunami hazard on the coast
The fact sheet does not project shaking or tsunami effects for a future rupture; it documents what past ones left behind, including coastal land-level change, tsunami inundation, landslides and liquefaction. Related USGS work fills in part of the picture. A study in the Proceedings of the National Academy of Sciences, described in a USGS article from May 5, 2025, found that a magnitude 8 or greater Cascadia earthquake could cause 0.5 to 2 meters, or 1.6 to 6.6 feet, of sudden land subsidence. The sinking would more than double the number of people, buildings and roads exposed to coastal flooding, with examples drawn from the Necanicum River, Yaquina Bay and Alsea Bay in Oregon and Humboldt Bay in California.
A USGS page titled Could It Happen Here? traces the same hazard to Japanese historical accounts of a tsunami generated by a local earthquake “close to magnitude 9,” and says geologic evidence of submerged vegetation points to large or great earthquakes of magnitude 8 to 9 about every 500 years on average. Great ruptures there would most likely produce local and possibly ocean-wide tsunamis. Against that record, a 15 percent chance over 50 years leaves an 85 percent chance that no full-margin rupture occurs in the window, and the five-point gap between the fact sheet’s 15 and 10 percent figures is the part of the estimate that depends on counting the years since 1700.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
Help yourself and the planet: Get The Cool Down →
We may earn a commission if you sign up.
More from Morning Overview
- The NSA says three phone features should be off whenever you aren’t using them
- Hurricane Hunter radar shows four warning signs that a tropical cyclone is about to strengthen, a University of Miami study found
- Tropical Storm Rachel is dumping up to 12 inches on four Mexican states on its way to major hurricane strength
- The FTC says Lens.com doubled the price shoppers saw in Google ads
