Fault finding may enhance quake predictions
By Sally Harris
Spectrum Volume 20 Issue 11 - November 6, 1997
A Virginia Tech geological-sciences professor and two colleagues from Rice University have discovered a fact about the San Andreas fault that may help in our understanding of earthquake hazards in California and other areas.
John Hole, assistant professor of geological sciences at Virginia Tech, said seismic reflection and refraction surveys of the deep crust show that the San Andreas fault goes straight through the crust and cuts through the Moho, the boundary between the crust and mantle of the Earth, instead of turning in the crust to connect with two other parallel faults in the area. The discovery, which helped discount recent suggestions that the faults linked at shallow depths, affects the predictions of the stress that builds up between the faults and the resultant assessments of earthquake hazards. The discovery was published October 24 in Science.
According to Hole, who did the study with Timothy J. Henstock, leader of the study, and Alan Levander of Rice's Department of Geology and Geophysics, the San Andreas fault is the major boundary between the Pacific and North American geological plates of the Earth. Seismic surveys help determine the shape and composition of the structures in the deep crust.
Hole and his colleagues did their surveys in northernmost California at the Mendocino Triple Junction, where the Pacific, North American, and Gorda plates come together. The Triple Junction is migrating northward parallel to the coast, Hole said. As it goes, it lengthens the North American-Pacific plate boundary, which is the San Andreas fault. The rock boundaries such as the Moho are at different heights, or different rocks are juxtaposed on either side of the fault, not because one side has moved up or down, but because the rocks are moving northward so that those that are on one fault are not the same ones that formerly were on that side of the fault. The study done by Hole and his colleagues found such offsets at lower crustal and Moho depths, proving that the fault goes that deep.
By setting up vibrations--explosions and compressed air pops--in the area and listening to them at hundreds of kilometers, Hole and his colleagues can tell what is inside the Earth on each side of the San Andreas fault down to 25-30 kilometers. The process is roughly analogous to ultrasound techniques that allow doctors to view the heart beating inside the body, Hole said, except that the much larger process used on Earth sends out very low-frequency vibrations that cannot be heard.
The resulting images show changes in the physical properties of the rocks, and, from those images, the geologists can infer the types of rocks present and see boundaries between the rock types.
The faults in the area being studied are very young geologically speaking--about two million years old. Learning how it is growing northward may also tell scientists how it grew past the San Francisco area, Hole said.
Questions the geologic community has been asking include how deep the faults go and whether the San Andreas fault goes straight down or turns in the crust to link to other faults. If the fault turned and linked with others at shallow depth, the stresses that build up would be different than those created if it goes straight down and would affect earthquake hazards, Hole said.
The ultimate goal of the study is to understand where the rock boundaries are, the types of rock, and the way stresses build up so that scientists can better understand earthquake hazards.
The study by Hole, Henstock, and Levander provided the first strong evidence that the San Andreas fault goes straight through the crust 25-30 kilometers and cuts through the Moho without connecting with other faults. That fact changes the view of the buildup of stresses in the other faults and improves our understanding of the buildup of stresses in neighboring faults and the resulting earthquake hazards there. "Because the links are deeper, the effects of motions on one fault are less directly linked to neighboring faults," Hole said. "Sometimes that can actually cause a greater effect, depending on geometry and other factors. But knowing the geometry, we can better model the effects."