BY DAWN LEVY
Move over, San Andreas. There's a new fault on the block. At the meeting of the American Geophysical Union in December, Stanford scientists presented evidence that geophysical forces are creating new faults in the Mojave Desert. What's more, the faults ruptured in patterns that puzzled scientists, who expected earthquakes along the established north-west and east-west lines of mapped, mature faults. Instead, they saw ruptures both along old faults and along new, previously unrecognized faults that ran north-south and north-east and cut across many old faults. Some of the ruptures broke the Earth's surface; many did not.
"What is really striking to us about these earthquakes is that they rupture faults with multiple orientations," says Greg Beroza, associate professor of geophysics.
In the past 67 years, seven moderate-to-large quakes have traced a 100-kilometer (62-mile) line through the Mojave. Their magnitudes ranged from 5.4 to 7.4. The largest was the 1992 Landers earthquake; the latest was the Hector Mine quake in October 1999.
"It has been a misconception that faults in the Mojave could not produce such large earthquakes," says Hagai Ron, visiting professor of geophysics from the Geophysical Institute of Israel and Hebrew University of Jerusalem.
Faulting at Hector Mine, which is near Twentynine Palms, Calif., occurred entirely within the U.S. Marine Corps Air Ground Combat Center. While most Mojave faults traverse similarly sparsely populated areas, at least six major faults cross California's well-traveled Interstate 40. The new faults cut across old ones that have rotated into orientations that make slippage difficult.
Quakes happen when faults slip. In a single earthquake, more than one fault may rupture, but the ruptures may not manifest themselves along a straight line. "The different orientations of faults are clear from the surface rupture patterns and from the aftershocks, and that's what we're trying to understand," Beroza says.
Ron, Beroza and geophysics Professor Amos Nur have developed a simple model that invokes fault-block rotation to explain nonlinear rupture patterns after a single earthquake. The model may explain the behavior of Mojave earthquakes in the past few million years.
The Stanford geophysicists employ Coulomb friction theory to predict the orientation of faults that form in the Earth's crust in response to stress. The orientation of a fault depends on the material properties and the relative magnitude of stress. The theory predicts, and experiments have shown, that faults usually develop at an angle of 30 degrees relative to the direction of the stress. Therefore, Ron says, only stress acting at certain angles can activate faults.
In the simplest analogy, think of the Earth's surface as a net: Tug on one corner, and all points in the net move. A fault slips at point A, and points B through Z move in response.
Now imagine a slightly more complicated analogy to demonstrate how geophysical forces change as land shifts. Just like a row of books will start to lean when the bookends that support it are removed, fault blocks of the Earth's crust can rotate and deform. And just like a row of books that slides too far to one side, blocks of crust can eventually stop slipping and lock up.
For further crustal deformation, a new set of faults must develop. "Formation of new faults is not a well-accepted concept," admits Ron. "But things have to be in equilibrium before and after an earthquake."
"We seem to be the only group that is seeking to explain the pattern of faulting seen in the Mojave with a mechanical model," Beroza says.
"Adversaries say the Earth is heterogeneous and things are complicated," Ron adds. "But we believe stress is pretty much uniform over an area like the Mojave."
Old faults in the Mojave may be 5 million to 6 million years old, whereas new faults may be only 10,000 years old. Because of the mechanics of faulting, new faults require greater force to slip than old ones. (There's more friction on a new fault, so greater stress is needed to shear the rock and drive slippage.) But both fault sets can slip simultaneously when activity shifts from old to new faults, Ron says.
As faults rotate and age, they become unfavorably oriented relative to stress and therefore less able to accommodate slippage. Older faults do not need much slip to rotate them to the angle -- about 60 degrees -- where they "lock up," Ron says.
Can new faults eventually stop old faults dead in their tracks? "Ask us in 100,000 years," Beroza says.
Scientists use two key methods to prove a fault has rotated. First, they can locate sedimentary rock, which is assumed to have formed when sediment deposited horizontally, and observe tilts in enormous sections.
Second, they can take paleomagnetic measurements -- that is, measure the direction of the magnetic field in a rock. "The Earth is a sphere, so we cannot tell if the crust is rotated relative to a point without an absolute frame of reference," explains Ron. "Our absolute frame of reference is the Earth's magnetic field because it always runs north to south. When rock is being formed from lava or even sediments, it contains magnetic grains. When lava cools or sediment deposits, the grains acquire the magnetic properties of the field and align parallel to the Earth's magnetic field, which is always north-south." In essence, Mother Nature has implanted tiny compass magnets in the rock.
"We can collect samples from any rock containing enough ferromagnetic material and measure the direction of these frozen compasses," Ron says. "This direction tells us how much a block of the Earth's crust has been rotated and in what sense -- say, clockwise or counterclockwise."
This technique revealed that older east-west faults in the Mojave have rotated clockwise 40 to 50 degrees, Ron says. This rotation effectively locks them up and sets the stage for the formation of new faults.
"Fault ruptures tell us about how faults work," Beroza says. But the lessons he, Ron and Nur unearth may extend well beyond the Mojave. New faults are forming around the world, and maps of surface faults alone may not provide enough information for scientists and engineers to adequately assess earthquake hazards. They will have to examine evolving geophysical forces to obtain a truer, three-dimensional picture of seismic risk.