The geological rupture responsible for the devastating magnitude-7.5 earthquake that struck Palu, Indonesia, in September 2018 ripped through Earth’s crust at rare high speed, two teams of scientists reported this week. This “supershear” behavior likely intensified the shaking in the quake, which triggered a tsunami and killed more than 2000 people. And the setting, on a fault not expected to sustain such a rupture, raises fears that far more regions could be at risk of high-speed quakes than once thought.
The Palu earthquake took place on a strike-slip fault, where two blocks of continental crust slide past each other laterally. From the start it stood out as unusual, says Lingsen Meng, a seismologist at the University of California, Los Angeles, and a co-author of one of the new papers, which appear in Nature Geoscience. Its shaking, for its magnitude, seemed especially powerful, causing widespread soil liquefaction and landslides. And, as satellite imagery rolled in, it became clear that the rupture had traveled some 150 kilometers, despite lasting only 35 seconds. “This was a very fast earthquake,” says Elizabeth Madden, a geophysicist at the University of Brasília.
Like rips in a piece of paper, earthquake ruptures don’t happen all at once. A rupture typically unzips a fault at a uniform rate of about 3 kilometers per second, below the speed of an earthquake’s damaging side-to-side waves, called shear waves, which spread out from the rupture tip.
Geology seemed to impose the speed limit: The rupture point chews up energy as it pulverizes rock, and seismologists thought a supershear rupture would consume too much energy to be sustained.
Whereas most earthquakes are content to obey this speed limit, scientists have clocked a handful that broke the supershear barrier, beginning with two 1999 earthquakes in Turkey. Meng says the ground shaking is generally much stronger in these cases. That’s because, as the rupture gathers speed, the earthquake shear waves begin to overlap, increasing in strength like overlapping waves in the ocean.
These quakes all seemed to take place on long, linear, and smooth strike-slip faults—geological runways where the sliding allows the rupture to gather speed and leap past the forbidden zone. But the Palu quake broke that rule. Meng and his co-authors tracked the speed of the rupture using variations in the arrival times of seismic waves at a dense array of sensors in Australia. They also analyzed satellite radar observations of the Palu region before and after the earthquake to learn how the rupture displaced the ground—a clue to the fault’s geometry. Rather than being a straight runway, the fault had big kinks. Yet the rupture still went supershear, traveling more than a kilometer per second faster than a typical earthquake.
The Palu quake holds other puzzles, Meng adds. Although it traveled at high speeds, it did not go quite as fast as previous supershear earthquakes, which typically run as fast as their leading pressure waves, at 6 kilometers per second. One factor may be the age of the Palu fault; it has likely hosted thousands of earthquakes, leaving shattered rocks that slowed the rupture’s progression. Also, rather than gradually building up speed like earlier supershear quakes, this one hit top speed immediately, like a jet going supersonic at takeoff. “Even in these complicated and rough faults, it can go supershear and it can go supershear right away,” Meng says.
A second study, using only satellite imagery of one segment of the rupture, supports the notion that the Palu earthquake went supershear. “We were immediately struck by the sharpness of the rupture at the surface,” says Anne Socquet, the study’s lead author and a geophysicist at the University of Grenoble in France. The ground seemed to slip almost seamlessly north and south, with little vertical motion, and the quake had no aftershocks—features consistent with past supershear earthquakes. Unpublished 3D computer models of the earthquake, designed to diagnose the tsunami’s cause, suggest only a supershear event can explain these observations, Madden adds.
Martin Vallée, a seismologist at the Institute of Earth Physics in Paris, says the evidence is convincing—and disturbing. By showing that even tortuous faults can break the speed limit, the finding means “it is difficult to exclude supershear behavior on most faults,” he says. Such quakes are still uncommon. But Meng says hazard assessments for strike-slip faults worldwide now need to reckon with the chance of intense supershear shaking.
As a resident of Los Angeles, perched above the San Andreas, perhaps the world’s most famous strike-slip fault, the threat is personal to Meng. “I don’t want to say that the San Andreas would go like that,” he says, but some believe it may have gone supershear in the past—in the 1906 earthquake that leveled San Francisco, California. “There’s definitely a possibility for this to happen.”