By Alexander Stone
Beneath 5,150 kilometres of rock and molten iron, something extraordinary is happening. Earth’s inner core – a solid ball of iron and nickel roughly the size of Pluto – appears to have stopped spinning faster than the rest of the planet. Around 2009, the differential rotation that scientists had spent decades detecting simply vanished from the seismic record. Then it came back, but in the wrong direction.
The discovery, first published in Nature Geoscience in January 2023 by Yi Yang and Xiaodong Song of Peking University, has since been confirmed by a separate team at the University of Southern California, reignited a decades-old scientific feud, and raised a deceptively simple question: what happens when the engine at the centre of the world stops turning?
The answer, it turns out, is both less dramatic and more unsettling than Hollywood once imagined. The 2003 film The Core – in which a stalled inner core triggers catastrophic electromagnetic storms and a heroic nuclear-powered restart mission – remains fiction. But the real geophysics unfolding beneath our feet has proven stranger, more contested, and considerably more interesting than any screenplay.
The Seismic Evidence
The inner core was not discovered until 1936, when Danish seismologist Inge Lehmann analysed P-wave shadow zones from earthquakes and concluded that a solid body must exist within the liquid outer core – a revelation that reshaped our understanding of Earth’s deep interior. For six decades after that, its behaviour was largely assumed to be static. Then, in 1996, Song and Paul Richards of Columbia University’s Lamont-Doherty Earth Observatory published a landmark paper in Nature providing the first seismic evidence that the inner core rotated slightly faster than the mantle – a phenomenon called super-rotation.
The method was elegant in its simplicity. Earthquakes that recur in nearly identical locations – so-called repeating earthquakes or doublets – send seismic waves through the inner core along effectively the same path. If the core is rotating relative to the mantle, those waves sample slightly different material each time, producing detectable shifts in waveform arrival times.
By the early 2020s, Yang and Song had assembled a dataset of 121 repeating earthquakes from the South Sandwich Islands, a volcanic arc in the southern Atlantic, recorded between 1991 and 2023. Their analysis, published on 23 January 2023, showed that every seismic path previously exhibiting significant temporal changes had shown little to no change over the preceding decade. The inner core’s differential rotation had paused.
“We believe the inner core rotates, relative to the Earth’s surface, back and forth, like a swing,” Yang and Song told AFP. The period of that swing, they estimated, was approximately 70 years – with the next reversal point projected around 2040.
Confirmation and Complication
The Peking University findings might have remained a solo voice in a crowded field were it not for a second team arriving at the same destination by a different route. In June 2024, John Vidale of the University of Southern California and Wei Wang of the Chinese Academy of Sciences published a companion study in Nature, also using South Sandwich Island earthquakes but extending the analysis to include data from Soviet nuclear tests conducted between 1971 and 1974, as well as repeated French and American nuclear tests.
Their conclusion was unequivocal. “When I first saw the seismograms that hinted at this change, I was stumped,” Vidale said. “But when we found two dozen more observations signalling the same pattern, the result was inescapable. The inner core had slowed down for the first time in many decades.”
The USC study added a crucial wrinkle: the core had not merely paused but begun backtracking – sub-rotating relative to the mantle. From 2003 to 2008, the inner core super-rotated; from 2008 to 2023, it reversed and moved two to three times more slowly back through the same path. The asymmetry of the forward and backward rates, Vidale argued, meant that existing models would need fundamental revision.
Then, in February 2025, Vidale’s team published a third paper in Nature Geoscience that further complicated the picture. Analysing the same dataset with additional earthquake pairs, they discovered that seismic waveforms at one receiver array in Yellowknife, Canada, exhibited anomalous changes between 2004 and 2008 that could not be explained by rotation alone. The inner core’s near surface, they concluded, was undergoing viscous deformation – changing shape as well as spinning. “I was staring at evidence the inner core is not solid,” Vidale said.
The 70-Year Question
The central dispute is not over whether the inner core has slowed – both major teams now agree on that point – but over what the oscillation period actually is and what drives it. Yang and Song propose a roughly 70-year cycle, with turning points in the early 1970s and around 2009. This periodicity, they argue, coincides with known variations in the length of day and the behaviour of Earth’s magnetic field, suggesting dynamic coupling between the planet’s innermost and outermost layers.
Others are not convinced. Hrvoje Tkalcic, a geophysicist at the Australian National University, has published research arguing that the inner core oscillates every 20 to 30 years, not 70. His analysis of paleomagnetic records suggests the cycle may be driven by magnetic field reversals rather than gravitational coupling. “These mathematical models are most likely all incorrect because they explain the observed data but are not required by the data,” Tkalcic said. “Therefore, the geophysical community will be divided about this finding and the topic will remain controversial.”
John Vidale, characteristically, takes a different view. He has argued elsewhere that seismic snapshots taken every decade are insufficient to map a 70-year cycle, and that the inner core may have moved significantly only between 2001 and 2013 before settling into relative stasis. The 2025 paper on near-surface deformation further muddies the waters: if the inner core’s surface is warping and reshaping on annual timescales, then what researchers have been interpreting as rotation may partly be topographical change.
What a Slowing Core Actually Means
The practical consequences of a slower-spinning inner core are real but minuscule. As the core decelerates, angular momentum transfers to the mantle, causing Earth’s rotation to speed up imperceptibly. The length of a day shortens – by roughly a thousandth of a second. “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,” Vidale said.
That millisecond matters for one reason: atomic clocks. Since 1972, Coordinated Universal Time (UTC) has been adjusted by the occasional insertion of a leap second to keep atomic time synchronised with Earth’s rotation. A 2024 study in Nature by Duncan Agnew of the Scripps Institution of Oceanography demonstrated that ice-melting-driven changes to Earth’s angular velocity had already postponed the need for a negative leap second – and that the core’s own deceleration was a contributing factor. If the inner core continues to slow or reverse, the timing infrastructure underpinning GPS, financial markets, and telecommunications networks may face new pressure.
The magnetic field story is more speculative – and more consequential than it might first appear. The inner core does not generate Earth’s magnetic field – that is the province of the liquid outer core’s convective dynamo. But the inner core’s solid structure stabilises those convective currents, and disruption could in principle weaken the geomagnetic shield. The Laschamps excursion, approximately 42,000 years ago, saw a temporary collapse of the magnetic field that correlated with heightened UV radiation, climate perturbation, and possible megafaunal extinctions. Whether a paused inner core could trigger anything comparable remains purely theoretical.
The Art of the Invisible
What makes this story culturally compelling is its fundamental mystery. The inner core is the largest solid object on Earth, yet no human being has ever seen it. Our understanding comes entirely from seismic waves – vibrations that propagate through thousands of kilometres of rock and arrive at the surface as wiggly lines on a screen. Scientists are performing sonar on a planet, using earthquakes as their pulses.
That act of inference has not gone unnoticed by artists. The British duo Semiconductor (Ruth Jarman and Joe Gerhardt) translate seismic data into immersive audiovisual works. Their piece Worlds in the Making appropriates the tools of volcanology to create landscapes hovering between science fact and fiction. Earthworks uses seismic recordings to drive computer-generated animations of geological formation – the data literally sculpts the image.
In Denmark, Seidlers Sensorium created EARTH CORE!, a sensory walk tracing the geological and biological life of the planet from its inner core to the atmosphere, with Inge Lehmann – the Danish seismologist who discovered the inner core – as guiding figure. The Australian artist Mel O’Callaghan’s Centre of the Centre descended to hydrothermal vents 2,500 metres below the Pacific, filming the elemental forces that may have seeded life itself.
These works share a preoccupation with the limits of human perception and the role of scientific instruments in extending it. The inner core’s slowdown is not merely a geophysical curiosity. It is a reminder that the planet’s most consequential processes operate on scales of time and space that exceed human experience by orders of magnitude.
Watching the Core Breathe
As 2040 approaches – the date Yang and Song projected for the next turning point – monitoring will intensify. The International Seismological Centre’s global network of broadband seismometers continues to expand. Machine-learning algorithms are being trained to decode seismic noise at resolutions unimaginable a decade ago. The European Space Agency’s Swarm satellites track real-time magnetic field variations from orbit, linking deep-interior dynamics to the magnetosphere above.
Vidale’s 2025 finding that the inner core’s surface is deforming – not just rotating but reshaping – may prove the most consequential revelation of all. If the boundary between the inner and outer core is more dynamic than previously assumed, the entire framework for modelling core dynamics will require rethinking.
“The mechanics are that the outer core is circulating and making a magnetic field, and so it’s kind of pulling the inner core back and forth,” Vidale told the Los Angeles Times. “We sort of think the outer core is stirring up the inner core, but the mantle’s trying to keep it aligned – maybe that’s why it’s oscillating.”
That image – of a planet’s interior caught in a slow-motion tug-of-war between electromagnetic and gravitational forces, playing out over decades – is not the stuff of disaster cinema. It is something quieter, stranger, and more enduring: the rhythm of a world that breathes on timescales we are only beginning to measure.
