Your turn
Each lab is one part of the video. Tap, slide and compare.
Each lab isolates one part of the story. The models are simplified to show the idea (USGS figures), not to forecast eruptions.
Tap and slide.
Fill the magma body, then drag the melt slider. The USGS says only about 5–15% is melted.
The shallow rhyolite body is mostly solid crystals with 5–15% melt (USGS). Eruptions are thought to need far more. Try it.
Slide to melt a little or a lot. Under Yellowstone, only a little is melted.
Tap each eruption. The cubes are sized by how much rock and ash came out.
Cube volume is proportional to erupted volume, relative to Mount St. Helens 1980 (USGS).
Tap a city or anywhere on the map to see the ash in the (very unlikely) worst case.
Tap a place for the modelled ash thickness band (Mastin et al. 2014). It assumes a caldera-forming eruption, the least likely scenario.
Tap a city. Lots of ash near the park, just a tiny bit far away.
Tap each event to see how often it happens.
Compare the frequency of hydrothermal explosions, lava flows and caldera-forming eruptions.
Tap a card. Small steam pops are the most likely.
Switch between a normal day and what a real warning would look like on the instruments.
Compare background seismicity and deformation with the unrest expected before a large eruption.
Switch between a calm day and a warning. Right now, it is calm.
Yellowstone is a park full of geysers. Hot water shoots up out of the ground. The water is heated by a giant volcano, hidden under the park.
Not a big lake of lava! It is mostly hot, hard crystals, packed together like a sponge. Only a little bit is melted.
Long, long ago, it had giant eruptions. Grey ash fell like dusty snow, and many animals survived. Scientists say a giant one is very, very unlikely. It may never happen again.
A little steam pop from a hot spring, or slow, slow lava oozing along inside the park.
It depends on the kind of eruption, and the likely kinds are small. If Yellowstone erupted in our lifetimes, it would most likely be a hydrothermal explosion (a blast of trapped steam and hot water) or a slow lava flow. Either would mostly affect the national park itself.42
The famous worst case is another giant, caldera-forming "supereruption". That would send scorching pyroclastic flows across nearby parts of Montana, Idaho and Wyoming, drop ash over much of the United States, and change the global climate for years to decades.5 But the U.S. Geological Survey (USGS) puts the chance of that in any given year at about 1 in 730,000, and says Yellowstone is not overdue.13
Yes. Yellowstone is one of the world's largest active volcanic systems.10 The park has more than 10,000 hydrothermal features, including more than 500 geysers, and all of them are heated by magma below.12 The heat flow from the ground here is about 30 times higher than is typical for the northern Rocky Mountains.10
The heat comes from a hotspot: a column of unusually hot rock rising from Earth's mantle. As the North American plate has slid southwest over it for about 16.5 million years, the hotspot has left a trail of old volcanic centers across southern Idaho's Snake River Plain, like a candle held under a moving sheet of paper.112
No, and this is the biggest myth about Yellowstone. Scientists map the underground using earthquake waves, a method called seismic tomography that works a bit like a medical CT scan.6 They've found two magma bodies:
| Upper (rhyolite) body | Lower (basalt) body | |
|---|---|---|
| Depth | about 5 to 17 km (3 to 10 mi) | about 20 to 50 km (12 to 30 mi) |
| Size | about 90 km long, 40 km wide | about 4.5 times bigger |
| How much is melted? | only about 5–15% | only about 2% |
So the upper body is mostly hot, solid crystals, with a little melted rock in the gaps between them. Think of a damp sponge, not a bathtub. The USGS says it is "unclear if there is even enough magma beneath the caldera to feed an eruption."63
A 2022 study that re-analyzed the earthquake data suggested the shallow body might hold closer to 16–20% melt on average. That's still well below the roughly 35–50% that scientists think an eruption would need, and the researchers stressed that the system is still mostly solid.15
Yellowstone has had three giant eruptions, about 2.1 million, 1.3 million and 631,000 years ago.2 Each one emptied so much magma from underground that the ground above collapsed into a huge crater called a caldera. The first caldera was up to 80 km (50 mi) long.2 The one visitors see today is about 30 by 45 miles.11
| Eruption | When | Size compared with Mount St. Helens (1980) |
|---|---|---|
| Huckleberry Ridge | 2.1 million years ago | about 6,000× larger (2,450 km³) |
| Mesa Falls | 1.3 million years ago | about 700× larger |
| Lava Creek | 631,000 years ago | about 2,500× larger |
| Mount St. Helens | May 18, 1980 | 1× (ash fell over 57,000 km²) |
Together, the three giant eruptions produced enough ash and lava to fill the Grand Canyon.278 Since the last one there have been about 80 smaller eruptions, mostly lava flows. The most recent was a lava flow on the Pitchstone Plateau about 70,000 years ago.2
In a giant eruption, the parts of Montana, Idaho and Wyoming closest to the park would be hit by pyroclastic flows: fast avalanches of scorching ash, rock and gas.5 Elsewhere the danger would be falling ash, thinning with distance.
USGS scientists used a computer model called Ash3d to simulate a modern supereruption. The ash cloud would spread out as a giant "umbrella", pushing ash more than 1,500 km even against the wind, so the ash would fall in a rough bull's-eye rather than a narrow strip downwind.9 The model found:9
The USGS expects such an eruption would have "enormous impacts on human activity (especially agricultural production) for many years."13 But the USGS notes the results are "far less than in most of these 'doomsday' scenarios" found online.9
Yes. A supereruption would be terrible for the region and hard on the whole country, but not the end of the United States. Gases like sulfur dioxide would rise high into the sky and cool the climate for several years, which would hurt harvests around the world.513 For comparison, the 1991 Mount Pinatubo eruption was about 1,000 times smaller than Yellowstone's biggest, and it cooled Earth by about 0.7 °C (1.3 °F) for around three years.13 And the USGS points out that even Yellowstone's past supereruptions did not cause extinctions.4
No. The USGS says plainly: "Yellowstone is not overdue for an eruption." Volcanoes don't follow schedules.3 Three giant eruptions give only two gaps between them (about 800,000 and 670,000 years). Two numbers are far too few to find a real pattern. Even if you average them anyway, you get about 725,000 years, which would put the "due date" roughly 100,000 years from now.314 Also, the USGS says scientists are not convinced another supereruption will ever happen at Yellowstone.3
For a giant, caldera-forming eruption, the USGS estimates the yearly chance at about 1 in 730,000, or 0.00014%. That's roughly the same as the chance of a 1-kilometer asteroid hitting Earth in a given year. The USGS adds that even this number is shaky, because it comes from averaging just two intervals.1
Smaller events are much more likely:414
| Event | How often | What it would do |
|---|---|---|
| Hydrothermal (steam) explosion | Small ones every few years; big ones every few thousand years | Usually a crater a few meters across; rarely hundreds of meters or more |
| Lava flow | About 80 eruptions since the last giant one; last one 70,000 years ago | Slow flows (at most a few hundred feet a day) mostly inside the park |
| Giant caldera-forming eruption | About 1 in 730,000 per year | Regional devastation, continent-wide ash, years of cooler climate |
The largest known hydrothermal explosion crater in the world is at Mary Bay on Yellowstone Lake: about 2.6 km (1.5 mi) wide, formed about 13,800 years ago.10 Steam explosions like these happen without any magma erupting. None of the large ones in the past 16,000 years was followed by a volcanic eruption.10 A rhyolite lava flow, like Yellowstone's most recent eruptions, would ooze slowly over months to years. That could cause fires and destroy roads in the park, but more distant areas would probably be largely unaffected.1110
Nobody knows when, or even whether, Yellowstone will have another giant eruption. The Yellowstone Volcano Observatory (YVO), a partnership of the USGS, Yellowstone National Park, universities and state geological surveys, watches the volcano around the clock.16 It monitors:
Before a large eruption, scientists expect intense earthquake swarms (with many quakes above magnitude 4 or 5) and rapid ground uplift, possibly tens of inches per year, across several parts of the volcano.9 Most scientists think the build-up to a catastrophic eruption would be detectable for weeks, and perhaps months to years.18 The USGS has found no sign that an eruption is imminent, and expects Yellowstone to stay eruption-free for the coming centuries.49
Want to know how volcanoes work in general? See our explainer How do volcanoes erupt? Yellowstone's swarms are small earthquakes, and its hotspot trail exists because tectonic plates move.
It depends on the kind of eruption. The most likely events are a hydrothermal (steam) explosion or a slow lava flow, which would mostly affect the park itself. A giant caldera-forming eruption, which is extremely unlikely, would send pyroclastic flows across nearby parts of Montana, Idaho and Wyoming, drop ash over much of the United States, and change the global climate for years to decades.
Yes. Even a worst-case supereruption would be a regional disaster, not the end of the country. Ash would be deepest near the park and only millimeters to centimeters thick on the East, West and Gulf coasts, according to a USGS model. Yellowstone's past giant eruptions did not cause extinctions.
In a giant eruption, pyroclastic flows would reach the parts of Montana, Idaho and Wyoming closest to Yellowstone. Ash would fall over much of the country, from decimeters to meters in the northern Rockies to centimeters in the northern Midwest. A small lava flow or steam explosion would mostly affect only the park.
No. Volcanoes do not erupt on a schedule. The USGS points out that the three giant eruptions give only two time intervals, too few to calculate a meaningful average, and even that average (about 725,000 years) would put the next one far in the future.
The USGS estimates the yearly chance of another caldera-forming eruption at about 1 in 730,000 (0.00014%), roughly the same as a 1-kilometer asteroid hitting Earth. Smaller hydrothermal explosions are far more common and happen every few years.
The last volcanic eruption was a rhyolite lava flow on the Pitchstone Plateau about 70,000 years ago. The last giant, caldera-forming eruption was about 631,000 years ago.
In a giant eruption, the ash cloud would spread outward like an umbrella in all directions, so southern Canada would likely get some ashfall, thinning with distance. Everyone would feel the cooler climate for a few years. Smaller, far more likely eruptions would not affect Canada.