The Why Files The Why Files -- whyfiles.org

Understanding quakes

POSTED 3 NOVEMBER 2005
The quake next door
Long ago, people who study earthquakes noticed that some parts of the planet seemed extraordinarily prone to shaking, and others were quite stable: California shakes, but not New York. Turkey shakes, but not the United Kingdom. But earthquakes also seem related in time: A big quake is usually followed by smaller aftershocks, which can last for a few years. And as we saw in the Andaman Sea, one big quake may be followed by another big quake on an adjacent section of the fault.

That's related in time and space, and it's a clue to what's going on underground.

San Francisco, 1906
Tall buildings withstood the 1906 shock much better than the housing in the foreground. Ground composition plays a major role in amplifying or dampening the seismic waves. Photo: USGS

Seismologists figured that somehow, through a process of "triggering," the first quake was setting up conditions that made the following quakes more likely. "Even 100 years ago people had noticed aftershocks, and thought maybe the crust was stressed by the mainshock," says Tom Parsons of the U.S. Geological Survey. "The idea of triggering goes back quite a ways, but with additional methods and computers the study has gotten more sophisticated."

In 2002, Parsons published a study (see "Global Observation..." in the bibliography) of whether one large earthquake could trigger a second large quake. He started by identifying all magnitude 7+ quakes between 1977 and 2000. He drew a 200-kilometer circle around the original quake, and counted magnitude 5.5 quakes within those circles. Then he calculated whether the original quake was more likely to increase or decrease the tectonic stress that led to the second quake.

Unsurprisingly, 60 percent of the follow-up quakes occurred in places where the stress increased. But why did 40 percent occur where the first quake apparently relieved the strain? Perhaps, Parsons says, the quakes were sparked by the ground motion of the original quake, which may have set up a small instability that grew over time. In the jargon, this is called "dynamic triggering," and it's been a "big source of uncertainty," Parsons admits. "Most models suggest you have to have [the follow-up earthquake] happen right when the dynamic waves pass through." But not all, he suggests. "Just as earthquakes damage buildings, you could look at damage to nearby fault zones. Would it change the contacts on both sides of the fault and alter the evolution of earthquakes?"

Friction governs whether the fault holds or slips, and, Parsons says, "there is a model of friction, where instead of faults being stuck all the time, they continuously slip very slowly. An earthquake happens after slip speeds pick up, becoming unstable. It may then be possible for a dynamically-triggered to happen after some delay while the slip evolution is completed."

At any rate, the study did confirm that follow-on quakes tend to vanish as the years pass. That, Parsons indicates, should help in the testy biz of understanding earthquake hazards.

But don't conclude that triggering is the Rosetta stone of seismology. Many other dangerous fault zones don't show a simple progression, says Thorne Lay. "It will pop off in one place, and jump to another, and then fill in the gaps. In Japan and South America, that means it's not a simple process that would allow us to pinpoint the next earthquake. Even in Turkey, we don't know when or how big it will be from the study of an adjacent earthquake. It's very complicated. We don't have a simple rule for predicting future events."

Time and Tide...
If an earthquake is able to trigger follow-on quakes further along the same fault, could smaller stresses and strains do the same thing? What about, say, the fear of an upcoming exam, or the agony of a flat tire on a busy highway? Oops, wrong train of thought...

Wooden structure towers over water during low tide, is almost engulfed during high tide.Historical tidal gauge at Anchorage, Alaska where the typical tide range is 35 feet. Photo: NOAA

Let's jump aboard the earthquake train...

Could the regular tidal movement of the ocean trigger a quake? Water seems weak, but as the old saw says, eventually it can erode rock. Could water trigger an earthquake?

For a century or more, seismologists have been looking at a possible correlation of tides and quakes, says Elizabeth Cochrane of the Scripps Institution of Oceanography in California. In a 2004 study (see "Earth Tides Can Trigger..." in the bibliography) she and colleagues found an association between strong tides and some types of earthquakes, which usually coincide with high or low tides.

"Water is heavy, and the tides are moving around a huge amount of weight on the Earth," says Cochran. "You have this large weight, pressing on a particular portion of the ground, then removing it."

Over time, this back-and-forth motion can have an impact, she says. "If you keep stretching a rubber-band, eventually it's going to break."

The researchers excluded quakes that were deeper than 40 kilometers in the crust, on the theory that deep rock is less affected by pressure on the surface. And because water exerts a vertical pressure on the sea floor, they looked only at thrust fault earthquakes -- those with some vertical fault movement. Previous researchers have found that tides have little or no impact on strike-slip quakes, where the movement is largely horizontal.

All tide up
When Cochran and Co. looked at the relationship between shallow thrust fault earthquakes from 1977 to 2002, the "overall correlation was quite small," she says. "This includes tides that are very low and very high in amplitude."

For the strongest tides, however, the researchers found a strong signal: Thrust fault earthquakes occurred at three times the background rate when the tide was highest or lowest.

Tides don't cause earthquakes, Cochran stresses, "they just change the timing. The earthquake would probably happen that day, but it was moved forward or back by the presence of water." So although the study may not directly advance earthquake prediction, it may help us understand how a slight change of stress can start the breakage in a stressed-out fault.When one section of fault quakes, it can set the stage for another quake further down the fault.

How do you know?
It's a question that bedevils seismologists as they try to map Earth's substructure. Their preferred method is to read the waves from distant earthquakes, and try to determine how fast the waves are being transmitted through the rock, and how they reflect off various underground structures.

Slower waves produce more ground shaking, explains Michael Ritzwoller, director of the Center for Imaging the Earth's Interior at the University of Colorado, and sedimentary rocks have slow transmission speeds, so in any given earthquake, sedimentary rocks will shake much more than igneous rocks. And since the Los Angeles Basin is a large sedimentary structure, a more exact picture of the rocks under the city might lead to smarter requirements for earthquake-proof construction or indicate where it's unsafe to locate, say, a chemical plant.

It sounds good on paper, but earthquakes are rare, and relying on them may force you to wait ages for data. Also, high-frequency waves don't travel far, so distant earthquakes usually don't supply these data-rich waves.

The result is a patchy picture of the subsurface. "The frustrating thing about earthquakes is that they don't occur everywhere, so your image may be tightly focused here, but very blurry there," Ritzwoller says. And while you can substitute large explosions for earthquakes, they are expensive, and generally unwelcome at the corner of Hollywood and Vine.

Earth hums: Recognize the tune?
Here's a novel solution: analyze the "seismic noise" that seismologists have long tried to ignore. Even when earthquakes are absent, these signals can be used "like a CAT scan" to probe the inside of the Earth, Ritzwoller says. The energy for these waves comes from ocean winds or storms that cause the ocean bottom to vibrate.

In a preliminary study using a large set of seismometers, Ritzwoller's results matched nicely with data from traditional sources (see "High-Resolution..." in the bibliography).

And there's one final advantage to the Earth-noise approach, Ritzwoller says. You don't even need to gather data. You can just scrounge it from your colleagues' trash. "These are waves seismologists always wanted to ignore," Ritzwoller says. "It's called Earth's hum, and it's always been considered a nuisance."

How deep can you drill?

backmore

Megan Anderson, project assistant; Terry Devitt, editor; S.V. Medaris, designer/illustrator; David Tenenbaum, feature writer; Amy Toburen, content development executive

©2021, University of Wisconsin, Board of Regents.