Reno, Nev. - Fault zone models do not always accurately portray earthquake potential and crustal movement in an area because researchers need to consider local earthquake history, according to a Penn State geophysicist.
When an earthquake occurs, the upper portion of the earth moves rapidly along the fault line, but the lower portion, because it is more plastic and stickier, takes time to catch up.
"We need to know the time duration since the last earthquake in order to interpret movement observations," says Dr. Kevin P. Furlong, professor of geosciences. "The apportionment of strain in a multiple fault system depends on how long ago an earthquake occurred in the area," he told attendees today (Nov. 14) at the annual meeting of the Geological Society of America in Reno, Nevada.
Furlong and a team of researchers from Penn State and the University of Miami, have been looking at movement along two fault lines in Baja California.
"While the boundary between the Pacific and North American plates in southern Baja California goes up the gulf of California toward the Salton Sea and the San Andreas Fault, the northern Baja Peninsula is cut by two faults running obliquely to the main plate boundary, " says Furlong.
About five million years ago, the Baja peninsula joined the Pacific plate, but while the transition is complete in the southern portion, the situation in the northern Baja is complicated. One fault, the San Miguel, runs in the direction of the plate motion as expected. The other fault, the Aqua Blanco, runs across the peninsula to the Pacific ocean.
The researchers use satellite-based global positioning systems (G.P.S.) to measure the motion of the crust to within millimeters. The motion along the faults in northern Baja is between 6 and 8 millimeters per year, but the motion is not equally distributed between the two faults. The offset on the San Miguel fault - the amount of past motion visible in the fault - is on the order of hundreds of yards, but the offset of the Agua Blanco fault is measured in miles. However, the Agua Blanco fault has little earthquake activity, while the San Miguel fault has most of the earthquakes.
"We are trying to understand the G.P.S. data and how the strain is partitioned between these two faults," says Furlong.
A variety of modeling techniques return similar regional activity, but have different results for apportioning the 8-millimeter annual motion. New modeling, using a three-dimensional finite element model that includes viscoelastic relaxation - the time it takes the ground under an earthquake to catch up with the surface - shows 25 percent of the strain on the Agua Blanco fault and 75 percent of the strain on the San Miguel fault.
The most recent earthquakes occurred on the San Miguel fault in 1956. If these measurements were done before that earthquake, the strain on both faults might have been equal. If they had been done only 10 years after the earthquake, the apportionment would be different still.
Because the Agua Blanco fault is oriented in the wrong direction relative to plate motion, its large displacementas compared to the San Miguel fault is puzzling. The researchers suggest that perhaps the Agua Blanco fault is an older fault and has had more time to move, or perhaps the Agua Blanco fault is a weaker fault and easier to move.
The researchers hope that additional observations and improved modeling will answer these questions, because the system has implications for earthquake hazard prediction. The Agua Blanco fault links up with an unnamed fault in the Pacific ocean, but the San Miguel fault aligns with the Vallecitos fault and may link with faults in the San Diego area. The San Miguel fault is an active earthquake fault and the way strain is partitioned between the two faults could impact future earthquakes in southern California.
Researchers working on the project include Furlong; Rocco Malservisi, Penn State graduate student in geosciences and Tim Dixon, professor and Julien Decaix, graduate student at the University of Miami.
Materials provided by Penn State. Note: Content may be edited for style and length.
Cite This Page: