Listening to the San Andreas

Discover what today's scientists are learning about the earthquakes to come
Peter Fish

Take a flight between Los Angeles and San Francisco and you see it about the time you finish your peanuts. Below you, running across central California, a scar as vivid as any caught on a plastic-surgery show like Dr. 90210 or Extreme Makeover. The San Andreas Fault.

One hundred years ago this month, the San Andreas changed history. The earthquake that struck the San Francisco Bay Area at 5:12 a.m. on April 18, 1906, shattered cities, helped create modern earthquake science, and established the image of California as a Marilyn Monroe-like place where seductive beauty joins with fatal instability.

A century later, earth scientists are uncovering the secrets of San Francisco's extreme makeover and the fault that caused it. "Like the San Andreas itself, our understanding of it has been locked," says Mary Lou Zoback, senior research scientist at the U.S. Geological Survey. "Now it is starting to move."

To understand California's next big earthquake, you need to understand the last one. Given that scientists have been studying the 1906 quake for a century, you might think we already know all about it. Yet even basic knowledge of where the quake was centered has shifted: not beneath Marin County, the hypothesis for decades, but farther south, offshore from San Francisco. And there remain many other gaps in our understanding.

This is why Jack Boatwright, a USGS geophysicist, found himself driving around the countryside north of San Francisco, looking for cemeteries near Sebastopol. He was trying to answer the question: How hard did the earth shake that April morning, and where did it shake most strongly?

Boatwright noticed that cemetery headstones were vulnerable in the 1906 quake. They toppled over. In many rural cemeteries, the fallen stones had never been replaced. By calculating the percentage of toppled versus upright pre-1906 headstones, Boatwright could determine how hard the ground shook.

So Boatwright spent months scouring Northern California graveyards. He hired his daughter, Phoebe, then 9, to help. "I paid her a dime for every headstone before 1906."

 

Based in part on his gravestone research, Boatwright and his colleagues have produced a new Shake Map for the 1906 earthquake, showing in vivid color the intensities of shaking that California experienced. The results are unsettling. The Sebastopol area, for example, shook far harder than anyone would have predicted, given its distance from the quake's epicenter.

You could look at this data two ways, Boatwright says. If you live right on top of the San Andreas, you may not be in as much trouble as you fear. "Or," he adds, "if you live 30 miles away from the fault, maybe you shouldn't feel too secure."

When that next big quake will strike is the question seismologists are invariably pestered with at cocktail parties. "Everybody always wants to know when it's going to occur," Boatwright says.

But for many earth scientists, quake prediction is a touchy subject. Some feel that specific prediction may be impossible ― that working on such research is as useful as studying 1-900 psychic hotlines.

We know more than we used to. Before the 1906 earthquake, scientists had no idea how often big quakes might strike the same stretch of fault. Every year? Every thousand years? The 1906 quake helped establish the elastic-rebound theory: Stresses build onto the fault, are released by an earthquake, then build again. Since the 1960s, it's been accepted that these stresses are created by parts of the earth's surface ― in California's case, the Pacific and North American plates ― sliding past each other.

But how fast are the stresses building? To know that, you need to know how fast the plates are moving, and in what direction. With the advent of global positioning systems, scientists can make these measurements with unprecedented accuracy.

Operated by the USGS and independent agencies such as UNAVCO (University Navstar Consortium), GPS stations resemble early prototypes of R2-D2 that never made it into Star Wars. USGS physical science technician Gary Hamilton has set up stations from the Farallon Islands off San Francisco to the High Sierra. Each one receives signals transmitted from U.S. Department of Defense satellites orbiting 12,000 miles above the earth. Scientists use the data to fix precise longitude, latitude, and elevation. With these, they can determine how fast the part of California sitting on the Pacific plate is moving compared to the part sitting on the North American plate: an average of 1.67 inches a year.

 

Seismologist Mary Lou Zoback sometimes gives talks where she outlines all the good things the San Andreas has brought California. Gold. Oil. And a lovely, rumpled landscape.

The area around Parkfield, in southern Monterey County, shows these benefits. Its oak-dotted hills are genuinely beautiful. But no place reveals the frustrations of earthquake science more clearly.

In the 1980s, seismologists noticed that the San Andreas Fault beneath Parkfield possesses an unusual trait: It seemed to produce neatly timed 6.0 earthquakes about 20 years apart. Parkfield became to seismologists what Burning Man is to millennial revelers: a pilgrimage site that generated wild hopes. An expensive array of monitoring equipment was set up, a long-term forecast made: Parkfield would feel the next 6.0 quake in 1988.

That quake finally arrived in 2004.

"It certainly showed that earthquake prediction is not going to be easy," says Mark Zoback, professor of geophysics at Stanford University and USGS scientist Mary Lou Zoback's husband.

Still, Mark Zoback adds, the Parkfield project generated much useful data. And now he is heading a new project in the same vicinity: the San Andreas Fault Observatory at Depth. For the first time, scientists are drilling into an active fault zone: 10,000 feet "into an earthquake laboratory."

SAFOD researchers have retrieved core samples from the fault, and they've inserted seismometers deep into it as well. "They will tell us what's happening inside the fault before, during, and after quakes," says Mark Zoback. "That's never been done before."

And then? Like a lot of earth scientists, Mark Zoback is aware of the difficulties inherent in forecasting quakes: the unlikelihood that anytime soon we can turn on Channel 2 and hear, "Partly cloudy tomorrow, with a 90 percent chance of a 4.5 earthquake." The work at Parkfield reinforces the question: Is quake prediction even possible? "We just don't know," he says. "But I'm open-minded about it.

"People are not stupid. They know that if we could predict quakes, we could be a hell of a lot safer. Just because a problem is hard doesn't mean you shouldn't work on it."

Info: Visit http://quake.wr.usgs.gov to find out more about the San Andreas Fault and other quake research.