Many of the world’s most dangerous earthquake faults are a silent menace: They have not ruptured in more than a century. To gauge the hazard they pose to buildings and people, geologists cannot rely on the record of recent strikes, captured by seismometers. Instead, they must figure out how the faults behaved in the past by looking for clues in the rocks themselves, including slickenlines, scour marks along the exposed rock face of a fault that can indicate how much it slipped in past earthquakes.
Now, researchers in New Zealand say slickenlines, when curved, can also reveal which end of a fault slipped first. “This is important information to know,” says Jean Paul Ampuero, a seismologist at the University of Côte d'Azur who was not involved in the work. Knowing how an earthquake ruptured in the past could help seismologists refine hazard assessments for cities, such as Los Angeles and Istanbul, that sit at the end of known faults. That’s because earthquakes beam their energy in the direction of rupture, Ampuero says. “If an earthquake is coming toward you, it’s coming to kick you in the face.”
Earthquakes don’t happen all at once. Rather, the slip between rocks begins at one spot on the face of the fault—the hypocenter—and travels along it, like a zipper being unzipped. As the rupture advances, the earthquake waves it generates pile up and intensify, like the siren of an approaching ambulance. Los Angeles lies at the northern terminus of the southern San Andreas fault, Ampuero notes. “If it breaks north, toward LA, that would be pretty bad.”
The researchers first noticed the curved slickenlines after the 7.8-magnitude Kaikōura earthquake, which struck New Zealand’s South Island in 2016. It was a chaotic event: The quake propagated from the southeast to the northeast and its energy jumped from fault to fault, causing dozens of them to slip and shake. One, the Kekerengu, crosses a series of canyons, creating more than a dozen exposures that revealed the marks of rock scraping against rock. Soon enough, the team noticed a consistently curved pattern in these striations on one side of the fault, says Jesse Kearse, a doctoral student in earthquake geology at Victoria University of Wellington and co-author of the work. “It was like a rainbow.”
More than 2 decades ago, Paul Spudich, a seismologist who died last year, had seen the same pattern in slickenlines from a Japanese earthquake. But no one had shown that such curves indicated anything about the rupture’s direction. Kearse showed his findings to Yoshihiro Kaneko, an earthquake dynamicist at GNS Science, New Zealand’s national geoscience research center. Kaneko thought his models could reproduce the motion that caused such an arch. When they fed his model with the Kekerengu data, the same shapes appeared. As in the real world, they formed on the side of the fault that had slipped toward the northeast.
The model also suggested how the arches form. The fault slip was mostly horizontal, but once the rupture reached the surface, no overlying rock constrained its slight upward motion. That allowed the side of the fault traveling with the rupture to bend slightly upward before leveling off. “It gets dragged off course by the seismic waves and makes an arch,” Kearse says.
That work, which Kearse and Kaneko published last year in Geology, covered just one earthquake. Kearse then plunged himself into the literature, unearthing 60 large historical earthquakes that had reached the surface and been documented by a geologist. Of those, one-third had curved slickenlines. Only eight of those 20 had the other constraints Kearse and Kaneko needed to test their model, like the earthquake’s hypocenter. But for all eight, the location of the curving slickenlines corresponded to the rupture direction predicted by their model, regardless of the earthquake’s magnitude or fault type, they reported in a paper published last month in the Journal of Geophysical Research. “It’s really compelling,” says Katherine Scharer, a paleoseismologist at the U.S. Geological Survey. “I’d love to see people go out after every rupture and see if they can document this.”
If similar slickenlines are discovered for older faults, that won’t immediately translate to better risk assessments, cautions Laura Wallace, a geodetic scientist also at GNS Science. Just because a fault ruptured in one direction in the past does not necessarily mean it will break in the same direction again. There are physical reasons to suggest that might be true, but the modern record is simply too short to say for sure. “It’s a huge question,” she says.
Slickenlines may hold the answer to that question, too, Scharer says. Not every fault has the right soft mudstones to preserve these lines, and even those that do likely only retain the last rupture. But under just the right conditions, a fault might capture multiple ruptures, she says, giving researchers a chance to look for sets of curved slickenlines that indicate the directions of multiple earthquakes. “It’s a precious moment of time being recorded.”
*Correction 25 September, 2 p.m.: An earlier version of this story misstated Jean Paul Ampuero’s affiliation.