Yellowstone’s floor keeps rising while scientists watch for the real hazard

The ground along the north rim of the Yellowstone caldera rose again in 2025, months after nearly six years without measurable change. Continuous GPS stations near Norris Geyser Basin recorded almost 2 centimeters of uplift between July and December, and satellite radar confirmed the same swelling pattern that scientists first mapped three decades earlier. None of it points toward an eruption. It is simply the latest entry in a deformation record that now spans more than a century, one the observatory’s lead scientist treats as routine even when the raw numbers sound dramatic.

Routine movement is exactly what keeps the U.S. Geological Survey’s Yellowstone Volcano Observatory watching the caldera floor as closely as it watches anything else in the park. The ground there does not sit still, and figuring out why it moves is how scientists decide which changes matter and which do not.

The ground under Norris Geyser Basin moved again

An area just south of Norris Geyser Basin, known to researchers as the Norris Uplift Anomaly, began rising in July 2025 after a period of near-total quiet, according to a January 2026 account from the Yellowstone Volcano Observatory. Michael Poland, the USGS geophysicist who serves as the observatory’s scientist-in-charge, wrote that continuous GPS stations northwest of the caldera began drifting away from the uplift zone that month, while semi-permanent stations installed for the summer recorded almost 2 centimeters of vertical rise between July and September. By the end of the year, horizontal displacement between the monitoring sites had reached about 1 centimeter.

Interferometric radar data collected by the Sentinel-1 satellite between October 2024 and October 2025 backed up the ground instruments, showing roughly 2 centimeters of uplift along the same stretch of the north caldera rim. The pattern closely mirrored an earlier episode, when the same patch of ground rose by about 12 centimeters between 1996 and 2004 before subsiding by roughly 7 centimeters over the following years. A smaller, shorter pulse of about 1 centimeter also came and went between 2020 and 2022.

The 2025 reactivation came with a modest uptick in earthquakes. Poland’s account notes that a swarm near the uplift zone produced more than 100 located earthquakes, mostly in November, with the largest reaching magnitude 2.7. Overall seismicity at Yellowstone stayed low for the year, with 1,113 earthquakes located against a typical annual range of 1,500 to 2,500, so the swarm stood out mainly by location rather than size.

A caldera that has risen and sunk for a century

Deformation at Yellowstone is close to constant. A 2021 explainer from the observatory notes that since 2015 the center of the caldera had been subsiding at a rate of about 2 to 3 centimeters a year, even as the Norris-area ground rose and fell on its own separate schedule a few miles away. Different patches of the caldera floor can move in opposite directions at the same time, driven by shifts in magma, the buildup and release of hydrothermal fluids, or simple fault motion far below the surface.

Scientists can track the movement in far finer detail than they could even a decade ago. The USGS page dedicated to the north rim uplift credits the improvement to denser GPS networks and satellite radar coverage, both of which can now pick up changes of a centimeter or less across a caldera roughly 30 miles wide. Poland has framed this sensitivity as part of the story: the 1996–2004 episode was one of the first deformation events at Yellowstone detected with this level of precision, and the tools have only improved since.

Why scientists rule out an imminent supereruption

The question that follows any Yellowstone headline is whether the volcano is “due.” The USGS answer, laid out in its own frequently-asked-questions page on the subject, is that the annual odds of a supereruption run at roughly 1 in 730,000. That figure comes from averaging the intervals between Yellowstone’s three known supereruptions, which struck 2.08 million, 1.3 million, and 631,000 years ago, for an average gap of about 725,000 years. The agency is candid about the limits of that math: with only two intervals to average, the number carries little statistical meaning, and volcanoes do not actually erupt on schedules.

The agency also points to a more basic obstacle to a repeat supereruption: there may not be enough melted rock left to fuel one. The rhyolite magma chamber beneath the caldera is estimated at just 5 to 15 percent molten, with the remainder solidified but still hot, and USGS scientists say it is unclear whether that qualifies as sufficient fuel. The most recent eruption at Yellowstone, for context, was a lava flow roughly 70,000 years ago, not a caldera-forming blast.

The hazards Yellowstone Volcano Observatory actually tracks

If a supereruption is a statistical long shot, the hazards the observatory spends its time preparing for look different. The same USGS guidance states that hydrothermal explosions or smaller lava flows are the most likely form any future activity would take. Hydrothermal explosions happen when shallow groundwater flashes into steam, sometimes violently enough to hurl rock and mud well beyond a geyser basin’s boundaries, and they can occur without any deep magma movement at all.

That distinction is why deformation like the Norris Uplift Anomaly gets tracked in monthly bulletins rather than emergency alerts. Poland’s write-up frames the 2025 uplift as evidence of an active hydrothermal and magmatic system working normally, not as a precursor. Before anything hazardous developed, he wrote, deformation rates would need to accelerate sharply — a change the continuous GPS network is built to catch well before it became visible on the surface.

Whether the current uplift is being driven by magma still accumulating roughly 14 kilometers down, the source blamed for the 1996–2004 episode, or by shallower hydrothermal fluids moving through cracked rock closer to the surface remains an open question. The observatory says it will keep watching the same GPS pairs that caught the signal in the first place, and report back through the observatory’s monthly updates.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview