Nearly 2,900 kilometers below the surface, just above Earth’s molten core, something has been quietly reshaping the planet’s deepest rock for millions of years. Now, for the first time, scientists have mapped that deformation on a global scale.
To do it, a team led by researchers at UC Berkeley assembled what may be the largest seismic dataset ever used to study this region — more than 16 million seismograms drawn from 24 data centers worldwide. What the data revealed points back to tectonic plates that disappeared from Earth’s surface long ago.
A global map of the deepest mantle
Jonathan Wolf and colleagues at UC Berkeley set out to study nearly 75% of the lowermost mantle — the layer sitting just above the core-mantle boundary, roughly 2,900 kilometers down. Seismograms were pulled from 24 data centers across the world, building a dataset of more than 16 million records. The resulting study, published in The Seismic Record, delivers the first global-scale view of how deformation is distributed at these extreme depths.
No single data center could have made this possible. The sheer scale of the dataset is what allowed the team to identify patterns that would otherwise stay invisible — spread across hundreds of kilometers of some of the most inaccessible rock on the planet.
Reading deformation through seismic anisotropy
When an earthquake strikes, it sends shear waves rippling through Earth’s interior. Those waves don’t all travel at the same speed — their pace varies depending on direction of travel and the properties of the surrounding material. That directional variation is called seismic anisotropy, and it’s one of the few tools available for detecting how deep rock has been stretched and distorted across geological time.
The team focused on wave phases that travel down through the mantle, pass into the outer core, and return back up. This path makes them particularly useful for sampling anisotropy across large lateral distances in the lowermost mantle.
Wolf has been direct about why this matters. In the upper mantle, deformation is well understood — largely driven by the drag of tectonic plates moving overhead, confirmed clearly by seismic anisotropy. The lowermost mantle has had no equivalent framework. “We don’t have any of this kind of large-scale understanding for flow in the lowermost mantle,” Wolf explained. “And that’s really what we want to get at.”
Subducted slabs leave their mark
The results were consistent in a notable way. Roughly two-thirds of the regions examined showed measurable anisotropy — and most of that deformation clustered in areas where ancient subducted slabs are thought to have settled after sinking through the mantle over millions of years.
Geodynamic simulations had long predicted this kind of connection between slab locations and deep mantle deformation. A prediction and an observation are different things, though, and this is the first time the link has been demonstrated at a global scale using seismic methods.
The mechanism behind the anisotropy isn’t fully resolved. One possibility is that the slabs carry a kind of “fossil” anisotropy — a structural memory from when they existed under shallower, different conditions. Wolf considers a more active explanation more likely: as the slabs sink and press against the core-mantle boundary, extreme heat and pressure reshape the minerals within them, generating a fresh anisotropic fabric, while the surrounding mantle material gets pushed and deformed in the process.
What the silent zones don’t mean
Not every region examined showed a detectable anisotropic signal. Wolf was careful to note, however, that silence in the data shouldn’t be read as evidence of undisturbed rock. In some areas, the signal may simply fall below what current methods can reliably detect — a limit of the technique, not necessarily a property of the mantle itself.
That caveat matters for interpretation. The map shows where deformation is detectable, not a complete census of where deformation exists or doesn’t.
Wolf described the full dataset as a “treasure trove” that researchers will keep returning to. The 16 million seismograms analyzed here represent a foundation, not a ceiling — future work could revisit the same data with different analytical approaches, or examine specific regions in far greater detail.
What comes next for deep Earth research
The broader ambition is to eventually map flow directions throughout the lowermost mantle — to understand not just where deformation has occurred, but how material is actually moving at these depths. Wolf has described that goal plainly: illuminating the lowermost mantle from many directions, at different lateral scales, until a coherent picture of global flow emerges.
That picture would carry implications well beyond academic geology. How material circulates at the base of the mantle influences heat flow out of the core, the long-term behavior of the magnetic field, and the slow cycling that ultimately drives plate tectonics at the surface. This first global map is an early but significant step toward understanding how Earth’s deep interior has been — and continues to be — shaped by its own ancient history.