Why do some rocks break so easily once an earthquake begins? In a giant quake, the fracture, where the two sides of the fault grind against each other, can extend dozens or hundreds of miles. The question has met several answers over the years.
Rock powder — ideal grease for earthquakes?
According to one theory, rocks get hot enough at the break to form a slippery layer of glassy rock along the fault. But that is not entirely satisfactory, says Ze’ev Reches, a professor of geoscience at the University of Oklahoma, because large earthquakes can form where the rocks are too cool to form glass.
“For some reason, friction seems to decline during a break, but what is the mechanism?” he asks.
In a laboratory study in this week’s Nature, Reches and David Lockner of the U.S. Geological Survey showed that a thin layer of rock powder that forms at the break causes a rapid drop in friction, which allows the break to spread further and faster down the fault. “The powder itself is a lubricant and it reduces the friction when it forms,” he says.
Haiti. January 12, 2010. Magnitude: 7.0.
San Francisco, CA. April 18, 1906. Magnitude: 7.8
Chile. February 27, 2010. Magnitude: 8.8, and March 5, 2010. Magnitude: 6.6.
Indonesia. September 30, 2009. Magnitude: 7.5.
Turkey. August 17, 1999. Magnitude: 7.6.
Rock ‘n rotate
To test rock samples, the researchers used a press that rotated one sample against another, producing a motion that was more representative of actual earthquakes, and also much longer and faster than previous researchers have studied. The study showed that a thin layer of rock powder can weaken the fault by at least 50 percent, Reches says.
The results concern how rock can slip once a fault breaks. It would be nice to know how the first rupture occurs, Reches says, but conditions in fault zones are so varied “that there is always a place where it’s significantly weaker, or is under a significantly higher load, so it starts moving. The question becomes, how far will this movement go?”
And the answer depends on how much friction remains in the broken portion, he says.
The rotary rock-grinder also showed that the powder, called gouge, ceases to lubricate within hours or days. “Everybody has seen the powder in faults and in experiments, but it was always taken for granted that the gouge does not change its properties,” Reches says. “What we have discovered is fundamentally different: The gouge has to be formed fresh, each time, to obtain this lubrication.”
The original powder is composed of grains that are “a few tens of nanometers across, but then because of adhesion between the grains it starts forming much larger clusters,” Reches says. The small grains can slip against each other, “but once they form these clusters, it takes a lot of energy to break them, so friction rises.”
“The internal workings of earthquake faults is one of the great unsolved problems of geophysics,” says Harold Tobin, a fellow fan of faults who is professor of geoscience at the University of Wisconsin-Madison. “Understanding the friction and mechanisms inside a fault, as it suddenly goes from hundreds of years of building up tremendous stress to rupturing in an earthquake, would help us understand why, where and when earthquakes occur. Experiments like the ones reported by Reches and Lockner are a key tool for getting at how earthquake faults slip.”
The study, Tobin says, is “a window on how an initial cracking turns into an earthquake. In my view, the study is not a game-changer in terms of our understanding of earthquake faults, but it does provide some solid data that will feed into better theories and models.”
Bearing it out
The study could help explain why the many tiny quakes that occur each day do not set off major quakes, Reches says. “In Oklahoma, we have magnitude 2 or 3 quakes but they don’t grow, because the conditions surrounding the break are not suitable. Why an earthquake occurs is not related to the initiation, but to the weakness that allows it to propagate.”
The phenomenon could also explain the “creeping section” of the San Andreas Fault, near Parkfield, California. Beyond both ends of the 120-mile section, the fault produces lethal, magnitude 8 quakes, Reches says, yet quakes in between release less than 10,000 times as much energy. “Although it’s on such a major active fault, the creeping section accommodates the motion in a very different mode. It might be that the rocks in this zone are not capable, once the motion starts, of creating the gouge that would lubricate it.”
With medium- and large-size earthquakes, Reches says, “the fundamental issue is, what is the mechanism of the weakening? What we have found is that once it starts moving, the formation of gouge makes it much weaker than before the movement started.”
– David J. Tenenbaum