It’s easy to forget that plate tectonics is a relatively young theory. Researchers have known for only a little more than half a century that Earth’s lithosphere is essentially a jigsaw puzzle of rocky plates sitting on top of a viscous mantle. The plates violently collide, rip apart, or sideswipe each other, usually in excruciatingly slow motion.
Crucial to the theory of plate tectonics is subduction, the process in which a plate of dense oceanic crust collides with a plate of less dense crust and sinks into the mantle. Since the 1960s, scientists have shown that subduction zones are particularly important for driving processes like mountain building and seismic hazards like earthquakes, tsunamis, and volcanoes. But less well understood is how the three-dimensional structure of subduction zones can influence these phenomena.
Two main things have held scientists back from studying the structure of subduction zones, according to Gavin Hayes, a geophysicist at the U.S. Geological Survey (USGS): a lack of data on the 3-D geometry of subducting plates (also known as slabs) and a lack of computing power and software that would allow researchers to visualize these zones in three dimensions. But that’s all beginning to change.
Researchers now are benefiting from new tools to peek inside Earth’s interior, and massive leaps in computing power are allowing scientists to constrain and visualize the 3-D structure of slabs. These new 3-D models are providing critical insights into long-standing gaps in geology’s unifying theory of plate tectonics, including why some mountains arise and some earthquakes occur hundreds of kilometers away from plate boundaries.
“It’s quite an exciting time for this type of science,” Hayes said.
From 2-D to Reality
“Traditionally, seismologists and geodynamicists have been focused on a two-dimensional or cross-sectional viewpoint when studying subduction zones,” said Kirstie Haynie, a USGS Mendenhall Research Fellow.
Even school-age children are familiar with the two-dimensional cross sections of one plate diving beneath another and disappearing into the mantle. Such images are good approximations of subduction zone dynamics, according to experts, and useful for, say, describing the tectonic setting of a recent earthquake. But they can’t convey much about the three-dimensional geometry of the underlying slabs.
There has been an explosion of seismic data over the past 3 decades, as seismic recording stations have proliferated and new tools like seismic tomography have emerged. These technologies have helped scientists like Hayes build databases like Slab2, which models the 3-D structure of slabs in subduction zones around the world.
“Seismic tomography is kind of like a CT [computerized tomography] scan of Earth’s interior,” Haynie said. The technique tracks the movement of seismic waves generated by earthquakes as they bounce off underground features, allowing researchers to reconstruct images of inner Earth. What all these new data show is that subduction zones are highly variable.
“What we’re seeing is that even in one subduction zone, the geometry of the downgoing plate varies—in its sense of curvature, in the inclination of the subduction zone, and also how deep the subducted plate goes,” said Margarete Jadamec, a geodynamicist and assistant professor at the University of Buffalo in New York.
Last year, Jadamec and colleagues fed data on slab morphologies into an open-source visualization software called the ShowEarthModel to create 3-D videos of every major subduction zone around the world.
“These virtual tours of the various subduction zones are a way for researchers to build a mental picture of what the subduction zone looks like in 3-D,” Jadamec said. “You realize with these movies that a 2-D representation is inadequate because we can actually see in three dimensions [that] the slab varies.”
The data points are tied to their geographic location on the virtual Earth, so viewers can see exactly where the slab geometries are changing. “It forces you to honor the data,” Jadamec said.
Armed with better data and more accurate renderings of what slabs and subduction zones look like, researchers can begin asking questions about how their geometry influences seismic hazards and processes like mantle flow. And evidence is mounting that the 3-D geometry of slabs has a significant impact on the geologic processes taking place at plate boundaries. This has helped Jadamec and others address some long-standing gaps in the original theory of plate tectonics.
A Mystery Solved
“Inherent in the theory of plate tectonics is that the plates are actually rigid, and the deformation is concentrated at the boundaries,” Jadamec said. But that’s not what we actually see on Earth. “What we find in many locations is that we have mountain building and earthquakes that occur far from the plate boundary, like 500 or 1,000 kilometers away,” she said.
Take Alaska, which sits atop the North American plate just where it meets the Pacific plate. The state has mountain ranges, volcanoes, and earthquakes in areas where the simple theory of plate tectonics wouldn’t have predicted them.
For instance, there are tall mountains near the Alaska-Aleutian subduction zone where the plates converge, but Denali, the tallest mountain peak in North America, is some 500 kilometers inland in the Central Alaska Range. Researchers long wondered why the deformation occurred so far from the plate boundary.
Alaska also has some unusual volcanoes. Volcanoes tend to form directly over subduction zones, but in some locations, including in Alaska, they pop up off to the side of subducting slabs.
Finally, Alaska is also the site of the second-largest earthquake ever recorded with modern seismometers, the magnitude 9.2 temblor known as the Great Alaska earthquake of 1964 or the Good Friday earthquake.
To better understand these anomalies, Jadamec created one of the first large-scale 3-D geodynamic simulations of the subduction zone in the region. This allowed her to study the area in southeastern Alaska where the slab comes to an end.
Until recently, researchers tended to ignore slab edges because it was too complex to numerically model the ways in which the mantle arcs around the slab edge, a process called toroidal flow that was first demonstrated in laboratory experiments.
In the first studies of 3-D flow dynamics, researchers used tanks filled with a gooey medium like honey to stand in for the mantle, pressed hard slabs into it, and tracked the flow of the viscous liquid around the edges. As computing resources advanced, researchers like Jadamec began using computational fluid dynamics simulations.
“One of the things that the 3-D slab models like the fluid dynamics experiments show is that in addition to toroidal flow, you get vertical upwelling zones,” Jadamec said. “These upwelling zones seem to spatially correlate with where we observe those anomalous volcanoes on Earth’s surface.”
Building on this work, Jadamec and Haynie went looking for an explanation for Denali in the slab geometry data and found one. “In south central Alaska, there’s a segment of the slab that’s horizontal or flat beneath the overriding plate,” Haynie said. “We think it’s coupled strongly to the overriding plate and that the flat slab is kind of pulling that overriding plate along the path of subduction.”
When Jadamec’s numerical model accounted for both the flat slab and the activity of a nearby strike-slip fault known as the Denali fault, it was able to accurately predict uplift exactly where Denali is located.
That coupling might also explain why the region is prone to such large earthquakes, according to Haynie, but she cautions that there are other factors at play in subduction zones besides the dip angle that could contribute to seismogenesis. Several other studies that look at slab geometries and earthquakes suggest that flat slabs have a role in generating large quakes, including a 2016 paper in Science that found that the biggest historic quakes tended to correlate with areas where 3-D subduction geometry models show that the subducting slab is broad and flat.
“I still think it’s an open question exactly what role subducting geometry plays in big earthquakes,” Hayes said. “But we’re beginning to build the data sets that allow us to better address these questions.”
From Earth to the Cloud
The major challenge now is that our understanding of the 3-D geometries of these underground slabs is “incomplete and constantly changing,” said Haynie, who is building a cloud-based version of the USGS Slab database that can be constantly updated, so that researchers working with them will always have the most up-to-date information feeding their models. “Every earthquake is a new data point.”
The USGS is focused on mitigating hazards. “We’re trying to use these geometries to inform things like our seismic hazard maps, and our understanding of seismic hazards more broadly, so that we can hopefully better mitigate [the damage from] these big earthquakes in the future,” Hayes said. But he notes that these 3-D geometries can also help answer questions about where volcanoes and mountain ranges form.
“These are questions that have been addressed before,” Hayes said, “but now that we’re getting these better data sets, it’s important that we revisit them and see [whether] some of the theories that we’ve thrown out in the past 50 years have held up.”
—Kate Wheeling (@KateWheeling), Freelance Writer