Nobody has ever seen Earth’s inner core, and nobody likely ever will. It sits about 5,100 kilometres down: a ball of solid iron and nickel about 2,440 kilometres across, sealed under the rocky mantle and the churning liquid outer core. So when a research team says the thing has changed direction, the obvious question is how they could possibly know.

The answer is mostly earthquakes. And what the readouts seem to show is genuinely odd. For decades the inner core spun slightly faster than the planet’s surface, running a little ahead of the ground under our feet. Around 2010 it appears to have slowed, and now it is lagging behind the surface for what a 2024 study argues is the first time in about forty years.

We should be upfront about a couple of things. We are not geophysicists, and this is one team’s reading of a dataset that other seismologists read differently. Treat what follows as reporting on a live argument, not a settled fact about the planet’s interior.

Why earthquakes are the main window

Because the core is unreachable, its motion has to be worked out from waves that pass through it. Researchers rely on a trick called the “repeating earthquake”: quakes that strike almost the same spot years apart and produce near-identical signals. Compare those signals carefully, and differences in the waves that travelled through the inner core can reveal changes deep inside Earth.

The 2024 team, led by USC’s John Vidale and Wei Wang of the Chinese Academy of Sciences, leaned hard on this. Their Nature paper compiled 143 pairs of repeating earthquakes, drawn from 121 events near the South Sandwich Islands recorded between 1991 and 2023. They also reached back further, using seismic records from Soviet nuclear tests in the early 1970s.

Even the specialists find the results counterintuitive. “When I first saw the seismograms that hinted at this change, I was stumped,” Vidale said. What convinced him was the number of records pointing the same way. “But when we found two dozen more observations signaling the same pattern, the result was inescapable. The inner core had slowed down for the first time in many decades,” he added.

What “slowed down” actually means

The signature the team looked for is subtle. When a wave’s shape changes and then reverts to an earlier form, the researchers interpret that reversal as the inner core returning to a position it occupied before, relative to the mantle above it. String enough of those matches together and you get a timeline: the core running ahead of the surface, slowing, then moving more slowly in the opposite direction relative to the mantle.

Keep in mind that this is not the core screeching to a halt or reversing into some catastrophe. The relative motion involved is tiny, and its most tangible effect is very small. The shift is associated with changes in the length of a day of roughly a millisecond. As Vidale put it, “It’s very hard to notice, on the order of a thousandth of a second, almost lost in the noise of the churning oceans and atmosphere.”

The reason any of this reaches beyond seismology is that the inner core is coupled to the machinery that makes a compass work. The solid inner core sits inside a convecting liquid outer core and is gravitationally coupled to the mantle. The motion of electrically conducting iron in the outer core generates Earth’s magnetic field, while the inner core can influence the pattern of that convection. A change in how the inner core turns is therefore a clue to processes operating deep inside the planet.

The argument is not over

The inner core’s motion has been contested for years, and it still is. An earlier 2023 study by Yi Yang and Xiaodong Song of Peking University reached a broadly similar conclusion, but explained it differently. They framed the slowdown as one turning point in a roughly seventy-year cycle, with a previous reversal in the early 1970s. Song’s account of the timing sounds close to Vidale’s: “We see strong evidence that the inner core has been rotating faster than the surface, [but] by around 2009 it nearly stopped,” he said. A shared endpoint, though, is not a shared theory.

Vidale is confident his group has the better account. “Other scientists have recently argued for similar and different models, but our latest study provides the most convincing resolution,” he said.

If anything, the picture has gotten more complicated. A 2025 Nature Geoscience study using the same general set of South Sandwich earthquake signals found evidence that the shallow inner core may also be changing shape. The authors concluded that the seismic variations are tentatively explained by both rotation and deformation near the boundary between the solid inner core and liquid outer core. That means some signals previously attributed to rotation may contain another process as well.

What we take from all this is less a firm conclusion than a sense of scale. A hidden iron ball thousands of kilometres below us appears to have drifted from running ahead of the mantle to trailing behind it, and the only trace at the surface is buried in seismic records and tiny variations in Earth’s rotation. Vidale suspects there is more to find. “The dance of the inner core might be even more lively than we know so far,” he said.