• Axial Seamount is the best-monitored submarine volcano in the world.
• It’s the most active undersea volcano closest to California.
• It could erupt by the end of the year.
A mysterious and highly active undersea volcano off the Pacific Coast could erupt by the end of this year, scientists say.
Nearly a mile deep and about 700 miles northwest of San Francisco, the volcano known as Axial Seamount is drawing increasing scrutiny from scientists who only discovered its existence in the 1980s.
Located in a darkened part of the northeast Pacific Ocean, the submarine volcano has erupted three times since its discovery — in 1998, 2011 and 2015 — according to Bill Chadwick, a research associate at Oregon State University and an expert on the volcano.
Fortunately for residents of California, Oregon and Washington, Axial Seamount doesn’t erupt explosively, so it poses zero risk of any tsunami.
“Mt. St. Helens, Mt. Rainier, Mt. Hood, Crater Lake — those kind of volcanoes have a lot more gas and are more explosive in general. The magma is more viscous,” Chadwick said. “Axial is more like the volcanoes in Hawaii and Iceland … less gas, the lava is very fluid, so the gas can get out without exploding.”
The destructive force of explosive eruptions is legendary: when Mt. Vesuvius blew in 79 AD, it wiped out the ancient Roman city of Pompeii; when Mt. St. Helens erupted in 1980, 57 people died; and when the Hunga Tonga-Hunga Haʻapai volcano in Tonga’s archipelago exploded in 2022 — a once-in-a-century event — the resulting tsunami, which reached a maximum height of 72 feet, caused damage across the Pacific Ocean and left at least six dead.
Axial Seamount, by contrast, is a volcano that, during eruptions, oozes lava — similar to the type of eruptions in Kilauea on the Big Island of Hawaii. As a result, Axial’s eruptions are not noticeable to people on land.
It’s a very different story underwater.
Heat plumes from the eruption will rise from the seafloor — perhaps half a mile — but won’t reach the surface, said William Wilcock, professor of oceanography at the University of Washington.
The outermost layer of the lava flow will almost immediately cool and form a crust, but the interior of the lava flow can remain molten for a time, Chadwick said. “In some places … the lava comes out slower and piles up, and then there’s all this heat that takes a long time to dissipate. And on those thick flows, microbial mats can grow, and it almost looks like snow over a landscape.”
Sea life can die if buried by the lava, which also risks destroying or damaging scientific equipment installed around the volcano to detect eruptions and earthquakes. But the eruption probably won’t affect sea life such as whales, which are “too close to the surface” to be bothered by the eruption, Wilcock said.
Also, eruptions at Axial Seamount aren’t expected to trigger a long-feared magnitude 9.0 earthquake on the Cascadia subduction zone. Such an earthquake would probably spawn a catastrophic tsunami for Washington, Oregon and California’s northernmost coastal counties. That’s because Axial Seamount is located too far away from that major fault.
Axial Seamount is one of countless volcanoes that are underwater. Scientists estimate that 80% of Earth’s volcanic output — magma and lava — occurs in the ocean.
Axial Seamount has drawn intense interest from scientists. It is now the best-monitored underwater volcano in the world.
The volcano is a prolific erupter in part because of its location, Chadwick said. Not only is it perched on a ridge where the Juan de Fuca and Pacific tectonic plates spread apart from each other — creating new seafloor in the process — but the volcano is also planted firmly above a geological “hot spot” — a region where plumes of superheated magma rise toward the Earth’s surface.
For Chadwick and other researchers, frequent eruptions offer the tantalizing opportunity to predict volcanic eruptions weeks to months in advance — something that’s very difficult to do with other volcanoes. (There’s also much less likelihood anyone will get mad if scientists get it wrong.)
“For a lot of volcanoes around the world, they sit around and are dormant for long periods of time, and then suddenly they get active. But this one is pretty active all the time, at least in the time period we’ve been studying it,” Chadwick said. “If it’s not erupting, it’s getting ready for the next one.”
Scientists know this because they’ve spotted a pattern.
“Between eruptions, the volcano slowly inflates — which means the seafloor rises. … And then during an eruption, it will, when the magma comes out, the volcano deflates and the seafloor drops down,” Wilcock said.
Eruptions, Chadwick said, are “like letting some air out of the balloon. And what we’ve seen is that it has inflated to a similar level each time when an eruption is triggered,” he said.
Chadwick and fellow scientist Scott Nooner predicted the volcano’s 2015 eruption seven months before it happened after they realized the seafloor was inflating quite quickly and linearly. That “made it easier to extrapolate into the future to get up to this threshold that it had reached before” eruption, Chadwick said.
But making predictions since then has been more challenging. Chadwick started making forecast windows in 2019, but around that time, the rate of inflation started slowing down, and by the summer of 2023, “it had almost stopped. So then it was like, ‘Who knows when it’s going to erupt?’”
But in late 2023, the seafloor slowly began inflating again. Since the start of 2024, “it’s been kind of cranking along at a pretty steady rate,” he said. He and Nooner, of the University of North Carolina at Wilmington, made the latest eruption prediction in July 2024 and posted it to their blog. Their forecast remains unchanged.
“At the rate of inflation it’s going, I expect it to erupt by the end of the year,” Chadwick said.
But based on seismic data, it’s not likely the volcano is about to erupt imminently. While scientists haven’t mastered predicting volcanic eruptions weeks or months ahead of time, they do a decent job of forecasting eruptions minutes to hours to days ahead of time, using clues like an increased frequency of earthquakes.
At this point, “we’re not at the high rate of seismicity that we saw before 2015,” Chadwick said. “It wouldn’t shock me if it erupted tomorrow, but I’m thinking that it’s not going to be anytime soon on the whole.”
He cautioned that his forecast still amounts to an experiment, albeit one that has become quite public. “I feel like it’s more honest that way, instead of doing it in retrospect,” Chadwick said in a presentation in November. The forecast started to garner attention after he gave a talk at the American Geophysical Union meeting in December.
On the bright side, he said, “there’s no problem of having a false alarm or being wrong,” because the predictions won’t affect people on land.
If the predictions are correct, “maybe there’s lessons that can be applied to other more hazardous volcanoes around the world,” Chadwick said. As it stands now, though, making forecasts for eruptions for many volcanoes on land “are just more complicated,” without having a “repeatable pattern like we’re seeing at this one offshore.”
Scientists elsewhere have looked at other ways to forecast undersea eruptions. Scientists began noticing a repeatable pattern in the rising temperature of hydrothermal vents at a volcano in the East Pacific and the timing of three eruptions in the same spot over the last three decades. “And it sort of worked,” Chadwick said.
Plenty of luck allowed scientists to photograph the eruption of the volcanic site known as “9 degrees 50 minutes North on the East Pacific Rise,” which was just the third time scientists had ever captured images of active undersea volcanism.
But Chadwick doubts researchers will be fortunate enough to videotape Axial Seamount’s eruption.
Although scientists will be alerted to it by the National Science Foundation-funded Ocean Observatories Initiative Regional Cabled Array — a sensor system operated by the University of Washington — getting there in time will be a challenge.
“You have to be in the right place at the right time to catch an eruption in action, because they don’t last very long. The ones at Axial probably last a week or a month,” Chadwick said.
And then there’s the difficulty of getting a ship and a remotely operated vehicle or submarine to capture the images. Such vessels are generally scheduled far in advance, perhaps a year or a year and a half out, and projects are tightly scheduled.
Chadwick last went to the volcano in 2024 and is expected to go out next in the summer of 2026. If his predictions are correct, Axial Seamount will have already erupted.