News From the Underground

By J. MADELEINE NASH LANDERS

"Talk about bad luck!" says Caltech geologist Brian Wernicke, squinting through a telescopic eyepiece at an aerial photo of Landers, California, a small town in the middle of the Mojave desert. "Wham! Right through this house. Wham! Right through that house. The funny thing is, there aren't that many houses out here."

In more ways than one, the earthquake that rumbled through this desolate region on June 28 was an ominous force. In a few fearsome seconds, it rerouted roads, realigned parking lots and reconfigured the landscape in countless capricious ways, miraculously taking only one life. Rather than rupture a single fault line, it swiped a 70-km (45-mile) diagonal slash through several, at one point heaving up a raw ridge of rock roughly the size and shape of a stegosaurus' spine.

For weeks afterward -- even this past week -- the region has been rocked by thousands of nerve-racking aftershocks, and the quake ignited mysterious swarms of smaller earthquakes in volcanic zones hundreds of kilometers away. But most alarming of all, this quake, the largest to hit Southern California in 40 years, appears to have substantially altered subterranean stress fields. In the process, it may have awakened a fitfully sleeping dragon -- the mighty San Andreas, the nation's biggest and most dangerous fault.

With a mixture of excitement and dread, scientists with the U.S. Geological Survey in Pasadena are rushing to augment an already extensive seismic network with portable instrumentation. "Before the San Andreas goes," reflects geologist Ken Hudnut somberly, "maybe we'll catch a precursor." A hot wind swoops across the desert as Hudnut retrieves a plastic box from under an oleander bush and pops the lid to reveal the small satellite receiver it shields from blowing sand. Nearby, a tripod-mounted antenna straddles a survey pipe like a spindly sentinel. Coded signals beamed down by orbiting ! satellites, Hudnut explains, serve to pinpoint the location of the pipe. The slightest shift in the pipe's position, and Hudnut will know the earth around it is on the move.

The southernmost section of the San Andreas has made scientists jumpy for some time now. Between 1948 and 1986, the region adjacent to the fault experienced only one earthquake of magnitude 5.8 or higher.* Since then there have been seven, including the Landers quake, which weighed in at an impressive 7.5. Moreover, this surge in seismicity appears to be occurring on a worrisome schedule. Excavations of old lake-bed sediments by Caltech paleoseismologist Kerry Sieh in the mid-1980s indicate that large earthquakes have roared through this section of the San Andreas at not quite 300-year intervals. The last such quake took place circa 1680. "It's just a gut feeling," ventures Sieh, who is 41 years old, "but I think I'll witness a great earthquake on the southern San Andreas in my lifetime."

About 1,300 km (800 miles) long, the San Andreas Fault system separates two sections of the earth's crust known as plates. Like giant rafts, these plates glide across an expanse of superheated rock, viscous as tar, that surrounds the planet's molten outer core. At the rate of nearly 5 cm (2 in.) a year, the Pacific plate to the west of the San Andreas is slowly pushing north, past the North American plate on the east. One possible result: 60 million or so years from now, a sliver of the California coast that includes the megalopolis of Los Angeles could become beachfront property in Alaska.

Getting there, however, will not be fun. The slip of the plates is not constant along the fault. The southern San Andreas bends like a river and splits into multiple branches. Because of this contortion, the Pacific and North American plates cannot slip in a straightforward way but must strain against each other like two sumo wrestlers. The battle of the plates has created numerous smaller fault lines along the San Andreas, giving the region the look of a smashed windshield. Over the millenniums, the Mojave shear zone to the east may offer a path of less resistance to the giant plates and replace the San Andreas as a new plate boundary, suggests geophysicist Amos Nur of Stanford.

Only four years ago, scientists gave the stuck plates along the southernmost section of the San Andreas a 40% chance of snapping sometime in the next 30 years. At the same time, they warned that a rupture of this part of the fault & could trigger earthquakes along neighboring segments, possibly as far west as San Bernardino and nearly as far north as Bakersfield. Result: the long-feared Big One -- an earthquake of magnitude 8, five times as powerful as Landers -- on the doorstep of the populous Los Angeles Basin. Now, in the seismic spoor of the Landers earthquake, scientists have found reason to suspect that the timetable for this disaster may have been fast-forwarded.

A fateful chain reaction, seismologists believe, started in April, when an earthquake of 6.3 magnitude rattled the vicinity of Palm Springs and Joshua Tree National Monument. On a map, the fault that was then broken looks like a shotgun taking dead aim at Landers, and in fact it was. Two months later, a minor earthquake started on a fault with no name. For a few seconds, this temblor rattled at a magnitude of 3. Suddenly, seismometer readings soared as the fracture unzipped a sequence of larger faults nearby. Then three hours after the Landers earthquake shivered to a stop, a 6.6 aftershock terrified the environs of Big Bear Lake, collapsing chimneys and toppling buildings.

Why Big Bear? In recent weeks research teams at the U.S. Geological Survey in Menlo Park have put this question to two different computer models. The results, while differing in detail, are strikingly similar. Before the effects of the Landers earthquake are taken into account, neither model flags the region around the Big Bear fault as particularly menacing. But as soon as scientists factor in the degree of ground movement and its direction, it pops up on their computer screens, color-coded red for danger.

Of late, the two teams have begun to use their computer models to peer into the future. What they see in these high-tech crystal balls is unsettling. "To relieve the stress Landers placed on it," says geophysicist Ross Stein, "the southern San Andreas would have to produce a 6.5-magnitude earthquake of its own."

The type of stress that has increased on the southern San Andreas is known as shear stress. It runs parallel to the fault, enhancing its tendency to slip. There is, however, another kind of stress, clamping stress, which retards slippage. It runs perpendicular to the fault, pinning the sides together like an invisible row of staples. "The situation we worry about most," says UCLA geophysicist David Jackson, "is when the shear stress increases and the clamping stress decreases. This is precisely what we think has happened."

| The Landers and Big Bear earthquakes cut through faults that form two sides of a triangle. When these faults fractured, the huge block of earth contained within the triangle shifted about a meter to the north, unclamping the San Andreas at the triangle's base. In deference to the menace posed by this singular geometry, Jackson calls the area the "Bermuda Triangle."

What everyone who lives in Southern California wants to know, of course, is not whether the southern San Andreas is going to slip, but when. To their frustration, scientists cannot answer that. The most careful calculations of stress transfer are based on the assumption that faults separate large blocks of earth, which stretch and compact in predictable ways. But the Southern California crust is so crisscrossed with faults that the material between them may behave more like sand. "Squeeze a block of wood," muses UCLA'S Jackson, "and it will become longer. But sand will behave in unforeseen ways."

In addition, scientists can only guess how much total stress accumulated along the southern San Andreas prior to the Landers earthquake. Geophysicist Geoffrey King of the Institut de Physique du Globe in Strasbourg, France, compares the predicament to trying to push a car uphill while blindfolded. Will the car move or not? "One person won't accomplish much," he observes, "but 10 people might. Our problem is that we don't know how many other people are already pushing on the car."

Indeed, scientists still do not know how much stress is required to start an earthquake in the first place. In laboratory experiments, explains University of Nevada, Reno, seismologist James Brune, two blocks of granite forced past each other generate a tremendous frictional heat. But earthquakes apparently do not. "Nature," says Brune, "has figured out an easier way of moving things around." After all, when carpet installers try to move a rug, they do not attempt to drag it all at once. "Instead," says Tom Heaton of the U.S. Geological Survey in Pasadena, "they put a little ripple in it. As the ripple moves from one end of the rug to the other, the rug moves with it."

What might cause such a ripple to spread across a fault remains a mystery. Numerous ideas have been suggested. Brune believes sliding rock physically deforms like tires squealing on pavement. In this case, what greases the skid is an invisible air pad that prevents the two surfaces from establishing frictional contact. Just last week, in a paper published by the science journal Nature, a team of researchers from the U.S. Geological Survey in Menlo Park offered an alternative possibility. Groundwater, they theorized, trapped under high pressure, might also serve to pry faults apart, allowing them to slip with a minimum expenditure of energy.

Yet another mechanism capable of inducing fracture has been suggested by the Landers earthquake. Because the quake triggered scores of sympathetic vibrations in volcanic and geothermal regions, some scientists have speculated that the Landers event shook underground magma chambers as though they were big cans of soda. The gas that fizzed forth could, in turn, have forced open a gap that eased the slip of surrounding rock. Whatever the mechanism, experts agree, it has only hastened the fracture of a fault zone that was already stressed up and ready to go.

What scientists fear is that the southern San Andreas has reached a similarly critical threshold. "If the Landers earthquake put a little stress on the San Andreas," exclaims Allan Lindh, chief seismologist of the U.S. Geological Survey, "then what about the accumulated stress of 300 years of plate motion?" For Lindh and other experts, the Landers quake and its resulting tremors are all too reminiscent of the increased seismic activity that preceded the great San Francisco blowout of 1906. "I mean," says Lindh, with a dramatic pause, "how much more on the edge of our chairs can we be?"

FOOTNOTE: *All magnitudes here are given on the moment magnitude scale, a more precise measure of earthquake energy that has largely replaced the Richter scale.