Bill Chadwick has forecast an Axial Seamount eruption correctly once before, and he is now working through his second attempt in real time. The research associate at Oregon State University, who has tracked the underwater volcano off the Oregon coast for years, revised his eruption window twice in less than a year, most recently pushing the estimate to sometime in mid-to-late 2026. That date is not a guess pulled from nowhere; it comes from comparing how much the seafloor above the volcano’s magma chamber has swelled against how much it swelled before the volcano’s last two eruptions.
Axial Seamount sits on the Juan de Fuca Ridge, a stretch of seafloor off the Pacific Northwest coast where two tectonic plates are slowly pulling apart, and it ranks as the most active submarine volcano in the northeast Pacific. It has erupted three times on record, in 1998, 2011, and 2015, each time preceded by a buildup researchers can now recognize and measure using an underwater sensor network wired directly into the volcano’s summit caldera.
A Volcano That Swells Before It Erupts
Axial’s eruptions follow a recognizable pattern: a period of elevated seismic activity paired with steady ground inflation as magma rises beneath the seafloor. Researchers track that inflation using instruments connected to the Regional Cabled Array, a network built from more than 660 miles of undersea cable linking upward of 140 separate instruments across the Juan de Fuca Ridge and the surrounding seafloor.
The last three eruptions each occurred once inflation reached a similar, though slightly rising, threshold, and Chadwick’s forecasting method rests on the assumption that the volcano will erupt again once it clears whatever the next threshold turns out to be. Inflation beneath Axial had slowed to nearly zero by the summer of 2023, after gradually rebuilding since the 2015 eruption, before both the inflation rate and background seismicity picked up again that fall, a shift Chadwick described as a fundamental change in the volcano’s magma supply.
The Forecast That Slipped From 2025 Into Late 2026
Renewed activity is what led Chadwick and colleagues, in a December 2024 presentation to the American Geophysical Union, to forecast an eruption within the following year. By late 2024, the seafloor above Axial had already climbed to about 95 percent of the inflation level recorded just before the 2015 eruption, close enough that a 2025 eruption looked plausible at the time.
Plausibility faded quickly. Inflation rates slowed again by the spring of 2025, and on October 27 of that year Chadwick updated the Axial Blog, the running public record he co-authors on the volcano’s behavior, to say the December 2024 forecast needed revisiting. Reaching the threshold the volcano cleared before its last eruption, he wrote, would simply take longer than the team had expected, putting the volcano on track for mid-to-late 2026 instead, a full year later than the window floated less than twelve months earlier.
A Threshold Set by the Volcano’s Own History
Ground level above the volcano’s magma chamber sat, as of the October update, roughly four inches higher than it had been just before the 2015 eruption, with perhaps another eight inches still to climb before the next one. The 2015 threshold itself sat about twelve inches higher than the one that preceded the 2011 eruption, a rising pattern Chadwick has compared to Krafla, an Icelandic volcano known for needing progressively more inflation before each successive eruption in a similar sequence.
That comparison is the basis for assuming Axial’s next threshold will again run somewhat higher than the last, rather than repeating 2015’s exact number. Chadwick has been candid that the method is closer to informed pattern recognition than hard physics, reading how past eruptions behaved and extrapolating forward, aware that inflation rates and thresholds can shift unpredictably at any point. Nothing about the pattern suggests thresholds will keep climbing forever, since the ridge periodically releases the compressive stress building in the surrounding crust as it spreads.
No Danger to People, Just a Model Being Tested
An eruption at Axial Seamount poses no threat to coastal communities; the volcano’s summit sits roughly a mile beneath the ocean surface, far from shipping lanes and well outside any populated area. What makes the eruption valuable to researchers is less the event itself than the chance to check a forecast against reality.
Starting in November 2025, Chadwick and colleagues began feeding real-time data from the seafloor sensors into a newer, physics-based model built to predict eruption timing more tightly than the pattern-based approach the Axial Blog has used until now. Chadwick has said the model’s results will not be released until after the volcano actually erupts, since only the eruption itself can show whether the new approach outperforms the simpler method that has guided the forecast so far.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- A granite sarcophagus surfaced in Egypt with its original lid still sealed
- The FBI tells phone owners to delete these toll-payment texts draining accounts nationwide
- Herbal supplements are landing Americans in the hospital with liver damage, doctors warn
- Long-term use of common heartburn pills is linked to kidney and dementia risk