Models show slabs of ocean crust (yellow) falling to Earth’s core.

Fabio Crameri

'Atlas of the Underworld' reveals oceans and mountains lost to Earth's history

Earth has a bad habit of erasing its own history. 

At intersections of tectonic plates worldwide, slabs of ocean crust dive into the mantle, part of the continuous cycle that not only drives the continents’ drift, but also fuels the volcanism that builds up island chains like Japan and mountains like the Andes. The disappearance of these slabs, called subduction, makes it difficult to reconstruct oceans as they existed hundreds of millions of years ago, as well as the mountains flanking them. “Every day, we’re losing geologic information from the face of the Earth,” says Jonny Wu, a geologist at the University of Houston in Texas. “It’s like losing pieces of broken glass as you’re trying to put it together again.”

But geoscientists have begun to pick up these pieces by peering into the mantle itself, using earthquake waves that pass through Earth’s interior to generate images resembling computerized tomography (CT) scans. In the past few years, improvements in these tomographic techniques have revealed many of these cold, thick slabs as they free fall in slow motion to their ultimate graveyard—heaps of rock sitting just above Earth’s molten core, 2900 kilometers below.

Now, the complete x-ray of Earth’s interior is coming into focus. Next month, at a meeting of the American Geophysical Union in San Francisco, California, a team of Dutch scientists will announce a catalog of 100 subducted plates, with information about their age, size, and related surface rock records, based on their own tomographic model and cross-checks with other published studies. “Step by step we went deeper and deeper, older and older,” says Douwe van Hinsbergen, a geologist at Utrecht University in the Netherlands, who led the project along with Utrecht geologists Douwe van der Meer and Wim Spakman.

Atlas of the underworld

A catalog of 100 slabs buried in Earth's mantle shows their size, age, and former positions near the surface.

(Map) J. You/Science; (Data) Van Hinsbergen/Van Der Meer/Spakman/Utrecht University

This “atlas of the underworld,” as they call it, holds the ghosts of past geography. By rewinding the clock and bringing these cataloged slabs back to the surface, scientists can figure out the sizes and locations of ancient oceans. Moreover, they can locate where the sinking slabs would have triggered melting, releasing blobs of magma that rose into the crust and drove volcanism. That has helped earth scientists pinpoint where ancient mountains rose and later eroded away, their traces visible only in unexplained rock records. “It’s a pretty exciting time to be able to pull all of these pieces together,” says Mathew Domeier, a tectonic modeler at the University of Oslo. 

That has only recently become possible, as the underlying technique, mantle tomography, is plagued with uncertainties. It relies on millions of seismic waves received by sensors scattered unevenly around the world. Waves with faster arrival times are assumed to have passed through the colder rock of subducted slabs. But seismometer coverage is patchy; earthquakes—the sources of the seismic waves—don’t occur everywhere; and the waves get fuzzier as they pass near the core or travel long distances. “Very often for regions that have the most interesting structures, you have the most uncertainty,” says Ved Lekic, a tomographer at the University of Maryland in College Park. 

Academic groups around the world use more than 20 models to interpret tomographic data, and their pictures of the mantle and its structures often conflict, says Grace Shephard, a postdoc at the University of Oslo. In the coming months, she will publish a comparison of 14 different models that will assess which slabs seem most likely to be real. Her results could cast doubt on some of the slabs in the Utrecht atlas. But the image of Earth’s interior is becoming more believable, thanks to improved computing power and such intercomparison projects.

By now the picture of lost plates is precise enough for scientists to try rewinding the clock, reconstructing vanished worlds. In earlier tomography, the plunging slabs looked like blobs in a lava lamp. But as the models have improved, the slabs in the upper mantle have been revealed to be stiff, straight curtains, says John Suppe, who heads the Center for Tectonics and Tomography at the University of Houston. The images make it clear that as they plunge, the 500-kilometer-thick slabs flex but don’t crumple—and that has made it easier for Suppe and others to unwind them. “We’re finding these plates unfold fairly easily, and they’re not that deformed,” Suppe says.

If this was all nonsense, it is really quite a coincidence.

Douwe van Hinsbergen, geologist, Utrecht University

These slab-driven reconstructions are calling into question plate movements inferred from ancient oceanic crust that was scraped off and preserved on the continents, Suppe says. “Almost everywhere we’ve looked at this,” Suppe says, “what we find in the mantle isn’t exactly what would be predicted.”

The reconstructions are also resurrecting mountains that had been lost to time. For example, in a study published several months ago, Wu and Suppe reconstructed the travels of 28 slabs to recreate the Philippine Sea as it was more than 50 million years ago. Beyond identifying what appears to be a previously unknown piece of ocean crust, they predicted that as one of their paleoplates plunged into the mantle, it threw up a large chain of volcanoes that eventually collided with Asia. That convulsive process could explain mysterious folded rocks in Japan and beneath the East China Sea.

Similarly, slabs beneath North America have helped bring that continent’s history of mountain building into clearer focus. By rewinding the clock for some of them, Karin Sigloch, a geophysicist at the University of Oxford in the United Kingdom, showed that North America’s western mountain chains, including the Rockies, likely formed between 200 million and 50 million years ago when several small plates were subducted beneath the continent, plastering multiple volcanic archipelagos against the landmass. 

Van Hinsbergen and his Utrecht peers hope their comprehensive atlas of slabs will make it possible to reconstruct a fuller picture of ancient geography. In 2012, they used slab tomography to constrain the longitude of volcanic island arcs that 200 million years ago dotted the ocean surrounding the Pangea supercontinent. Two years later they used their global model to estimate the number of subduction zones that would have been active over the past 250 million years, along with the amount of carbon dioxide (CO2) that subduction-related volcanoes would have emitted. The estimate closely matched geologic proxy records for atmospheric CO2 over the same period. And earlier this year, Van Hinsbergen published a study in Science Advances with Lydian Boschman, a graduate student, that identified several slabs that may have played a role in the birth of the Pacific Ocean. “We have done it,” Van Hinsbergen says. “If this was all nonsense, it is really quite a coincidence.”

Even with these new techniques, which Suppe together calls “slab tectonics,” the mantle’s memory of ocean slabs only stretches back 250 million years—the time it takes for one to fall to the bottom of the mantle and be fully recycled. Beyond that, Earth continues to cover its tracks.