On September 7, 2017, a magnitude 8.2 earthquake struck southern Mexico, killing dozens and injuring hundreds. While earthquakes are common enough in the region, this powerful event wasn’t any run-of-the-mill tremor.
That’s because part of the roughly 37-mile-thick tectonic plate responsible for the quake completely split apart, as revealed by a new study in Nature Geoscience. This event took place in a matter of tens of seconds, and it coincided with a gargantuan release of energy.
“If you think of it as a huge slab of glass, this rupture made a big, gaping crack,” says lead author Diego Melgar, an assistant professor of earthquake seismology at the University of Oregon. “All indications are that it has broken through the entire width of the thing.”
Such colossal fragmentation events have been observed before in a handful of places around the world, and all these epic earthquakes have one thing in common: No one really knows how they happen. This information gap matters, because huge populations from the western seaboard of the Americas to the eastern shores of Japan could be threatened by these enigmatic earthquakes.
For one thing, the deep quakes can induce strong shaking over a wide area that can level plenty of multistory buildings. One that took place beneath the Chilean town of Chillán in 1939, for example, killed at least 30,000 people. And when they happen near an ocean coastline, their destructive potential could be magnified.
“My real worry over these kinds of events is the tsunami,” Melgar says.
World’s most elusive earthquakes
Tectonic plates, also known as lithospheric slabs, are made up of the planet’s crust and the hot-but-solid upper mantle. They constantly move around Earth’s surface, either grinding side by side, crumpling up into one another and forming mountains, or descending under another plate in what is referred to as a subduction zone.
Along these various plate boundaries, you get earthquakes when friction generates stress that’s ultimately released. But quakes can also occur far from these plate boundaries, in the part of the slab that’s been pushed through a subduction zone and into the lower mantle. (Here’s what will happen when Earth’s tectonic plates grind to a halt.)
“If you bend an eraser, you can see the top half being extended and stretched, whereas the bottom bit is squashed and compressed,” Melgar notes. The same applies to these slabs. This bending can activate faults within the slab and trigger what are known as intraslab earthquakes.
Intraslab quakes are happening all the time at low to moderate magnitudes, often on faults involving side-to-side movement or the upward push of a block. On occasion, some incredibly energetic ones happen on so-called normal faults, where the movement of a chunk of rock follows gravity’s lead as it falls downward.
Melgar points to the 1933 Sanriku earthquake in Japan, which came in at a magnitude 8.5, as a good example of one of these intraslab normal quakes. Another would be the magnitude 7.8 Tarapaca earthquake in northern Chile in 2005. Sometimes, as in southern Mexico, the rupture can cut right through a slab. The same is thought to have happened beneath Iran in 2013 during a magnitude 7.7 tremor.
The hotel 'Ane Centro' was damaged after a 8.2 magnitude earthquake in Matias Romero, Oaxaca, Mexico.
Whether they feature this type of dramatic severance or not, these powerful quakes are inherently mysterious. Seismic surveys normally used to visualize tectonic movements can’t penetrate to such depths. The mapping of oceanic slabs is also in its infancy, and there’s not much high-resolution historical data to go on. That means geoscientists are currently scrambling for ways to best explain what’s going on.
Plate tectonics pandemonium
The new study’s geophysical measurements and models found that the Tehuantepec quake in Mexico was even more bizarre than any of the others. Normal faults can only rupture where the slab is being extended within the shallower segments. The Tehuantepec quake rupture, however, spread to even deeper parts of the slab that should be compressed.
This is potentially solvable. The paper suggests that the slab is being pulled down by its own weight so effectively that gravity is creating a major extensional force. This trumps the expected compressional forces, thereby allowing normal faulting to take place.
Far more problematic is the rupture’s staggering reach, which extended to a depth of around 47 miles. At this point, temperatures exceed 2,012°F, hot enough to permit the rocky slab to act more like a mushier plastic. A quake like Tehauntepec requires rock to be cooler and therefore harder, so it can break in a more brittle way.
Powerful normal fault earthquakes can take place in deep-ish parts of slabs, says study coauthor Emmanuel Garcia, a tectonics expert at Kyoto University. However, this only really applies to truly ancient tectonic plates that have had many millions of years to cool down, which makes them more prone to break in a brittle fashion.
The Tehauntepec quake involved the Cocos plate, which is a relatively young 25 million years old and is somewhat warmer than plenty of other tectonic plates. That, according to Melgar, makes the 2017 slab-splitting tremor “unheard of.”
Making a break for it
Part of the solution, according to Melgar’s team, may involve deep water. As the Cocos slab heads into the subduction zone under the North American plate, it bends and cracks. This creates normal faults, which take in seawater. As the slab passes into and through the subduction zone into the lower mantle, it warms up and dehydrates. This dehydration creates mechanical weaknesses and can cause brittle fracturing, creating small quakes or, perhaps, a huge one. The same theory was applied to the 2013 Iran and 2005 Chilean quakes.
The fact that the Cocos plate is younger and warmer could have created a “perfect storm” of events, suggests Stephen Hicks, a seismologist at the University of Southampton who was not involved in the new research. The plate’s relative warmth might mean that the vital dehydration process took place quicker, creating brittle conditions and faults early on that could eventually slip in a violent manner.
Melgar adds that when the oceanic Cocos plate first formed at a fiery mid-ocean trench, its cooling pattern created little hills and valleys in its rock. These imperfections may have eventually formed zones of weaknesses that could have generated the Tehuantepec earthquake, making this a story of destruction tens of millions of years in the making.
However, he notes, it still seems curious that brittle fracturing could take place so spectacularly at such hellishly hot depths. The slab could be oddly cold or made up of some strange rocks, he suggests, but both ideas go against what scientists expect conditions down there to be like.
Either way, figuring out the root cause of intraslab normal quakes is more than just an intellectual endeavour. Whether they are shallow or deep, these tremors can be powerful enough to suddenly shift any seafloor nearby, pushing vast quantities of water forward and creating tsunamis.
The Tehauntepec tremor took place on the landward side of the subduction zone, so the seafloor wasn’t deformed enough to create more than a 10-foot tsunami. By contrast, the 1933 Sanriku quake took place on the oceanward side of the subduction zone and created a devastating 66-foot tsunami.
When it comes to these strange, destructive earthquakes, “we don’t truly know what’s happening, to be honest,” says Hicks. But it’s clear that solving this titanic mystery could one day be a life-saver.