Jan. 19, 1998 Berkeley, CA -- For years geologists have tried to understand why the San Andreas Fault is so weak. In work supported by the U.S. Department of Energy's Office of Energy Research, geochemists Mack Kennedy, Yousif Kharaka, and their colleagues have found that part of the answer lies in the fault's surprisingly intricate connections with the Earth's mantle, deep underground.
The San Andreas, a classic strike-slip fault, marks the collisional boundary where the Pacific and North American plates meet. The forces at the boundary are compressive, yet fault failure is by shear, as the Pacific plate slides steadily if intermittently northward. Friction measurements in the laboratory on fault zone materials suggest that considerably more shear stress than is actually observed should be required for the fault to fail and the earth to move.
"The forces and movement of the fault should produce frictional heating," says Kennedy, a member of the Earth Sciences division of the Ernest Orlando Lawrence Berkeley National Laboratory, "but paradoxically, nobody's seen the expected heating in the vicinity of the fault. One possibility is that high-pressure fluids are acting as a sort of lubricant." Abnormally high pressures have been measured in rock pores at shallow depths, Kennedy says, "but to fully understand how the fault works it is extremely important to find out exactly what's down there."
Geologists including Mark Zoback of Stanford University have proposed drilling a deep hole right through the fault, three kilometers deep or more. "It occurred to us that if a bore hole encountered fluids, we would need to know where they came from. We set out to do chemical studies of fluids associated with the fault system, as well as measuring the ratios of helium isotopes," says Kennedy. "We located all the springs, seeps, and wells we could that showed evidence of deep-circulating fluids. We sampled them for carbon dioxide, hydrogen, noble gases, and so on. The fluid chemistry was in equilibrium with the local geology, as we'd expected, but in the course of this work, we found a helium-three signature in all the samples, which we did not expect."
Kennedy determines helium ratios using a a sophisticated gas-separation system and mass spectrometer, mounted in a truck trailer that can go on location when necessary. He found variable but comparatively high ratios of rare helium three (helium with only one neutron in its nucleus) to more common helium four (whose nucleus consists of two neutrons and two protons) in the San Andreas fluids, which proved to be telling clues to their origin.
Two competing models have sought to explain the origin of high-pressure fluids in fault zones. One, the Byerlee-Sleep and Blanpied model, or "closed box" model, suggests that local crustal fluids, including groundwater, are drawn into the fault zone in response to fault rupture and become trapped by mineral reactions; when the sealed fault zone compacts, the high fluid pressures required to weaken it are reestablished.
In the Rice model, by contrast, high fluid pressure in the fault is only the tip of a vertical "tongue" of high-pressure fluids originating in the mantle, 30 kilometers deep and deeper, that are focused into the fault zone by a root zone through the ductile base of the crust.
The Earth's atmosphere contains fewer than one and a half helium-three atoms for every million atoms of helium four. In crustal fluids, the ratio is even less -- only two hundredths of the ratio in air. But in mantle fluids, the ratio of helium three to helium four is about eight times greater than in the air. In fluids from the San Andreas Fault region, Kennedy and his colleagues found helium-three ratios that varied from over a tenth to as much as four times the ratio in air -- high ratios that were unrelated to the fluid chemistry in the local rocks.
"Some of this fluid could have come only from the mantle," says Kennedy. "The Rice model is at least partially correct."
The degree to which high-pressure mantle fluids contribute to the weakness of the San Andreas Fault, while large, remains indefinite, because Kennedy and his colleagues can't be sure if their sample fluids were tapped directly from the fault zone or from the adjacent crust. Meanwhile, the discoveries have raised interesting questions about the structure of the fault itself.
As fluids flow upward, helium three from the mantle is increasingly diluted by helium four produced from the steady radioactive decay of various elements in the crust. The ratio at a given site yields an estimate of how quickly the fluid reached that site from the mantle. The distribution of Kennedy's results leaves open the possibility that mantle fluid is flowing into the San Andreas Fault from great distances away.
"There may be a regional decollement that extends as far east as the Sierra Nevada -- maybe even under the Sierra," says Kennedy, noting the presence of soda springs near the crest of the Sierra which contain carbon dioxide that may have come from the mantle.
As for the nature of the mantle fluid, Kennedy says, "We don't know the chemistry, but it's likely to be rich in carbon dioxide and perhaps water under tremendous pressure" -- a mystery even a deep well won't answer in a straightforward way -- "but we'd really like to get fluids directly from the fault, to help us understand what makes the fault move the way it does. That's one of several good reasons to bore a deep well."
Kennedy is a member of Berkeley Lab's Center for Isotope Geochemistry (website at http://www-esd.lbl.gov/CIG/cig.html). He and his colleagues presented their results in an article in Science, 14 November 1997.
The Berkeley Lab is a U.S. Department of Energy national laboratory located in Berkeley, California. It conducts unclassified scientific research and is managed by the University of California.
Other social bookmarking and sharing tools:
The above story is reprinted from materials provided by Lawrence Berkeley National Laboratory.
Note: Materials may be edited for content and length. For further information, please contact the source cited above.
Note: If no author is given, the source is cited instead.