Å®ÉúСÊÓƵ

A substantial build-up of strain on a section of the San Andreas Fault south of San Francisco has been detected by an Australian-made instrument. The strain is accumulating at twice the rate measured by the same instrument a year before the Loma Prieta earthquake of October 1989. This earthquake, which was centred about 100 kilometres from the heart of San Francisco, caused fires in the Marina district of the city and caused sections of the Bay Bridge and the Nimitz Freeway in Oakland to collapse (This Week, 28 October 1989, 4 November 1989).

Strain in rock is the deformation that occurs when it is subjected to stress by the motion of tectonic plates – in this case the Pacific Plate sliding past the American Plate. ‘We are seeing a millimetre of change in length per year for every kilometre along this particular section of the fault,’ says Mike Gladwin of the University of Queensland. ‘That’s a lot of deformation.’

The build-up of strain is occurring near the town of San Juan Bautista, about 40 kilometres south of Loma Prieta. ‘(It) is consistent with suggestions of an increased probability of a moderate earthquake near San Juan Bautista,’ Gladwin and his colleagues, Ross Gwyther and Rhodes Hart, write in Nature (vol 356).

Gladwin believes that the strain that is building up has come from the 40-kilometre-long section of the San Andreas Fault that failed during the Loma Prieta earthquake. ‘Forces can’t disappear,’ says Gladwin. ‘They have to go somewhere.’

It is possible that the fault near San Juan Bautista can absorb the extra load, says Gladwin. But it is equally possible that the fault will fail again when the load reaches a certain critical threshold. ‘We can’t say that the new load will cause an earthquake,’ Gladwin says. ‘But we can say that without the load, it is improbable that an earthquake will happen. It is one of the conditions needed.’

Seismologists generally accept that sections of a fault interact with each other, and that when the strain is released along one section it can ‘load’ another. However, evidence to support this view has been scant because of the lack of instruments in the ground.

Gladwin has designed an instrument called a borehole tensor strain meter. Five of the instruments, including one near San Juan Bautista, have been operating in California since 1983. Each is set in eight tonnes of concrete in a borehole about 300 metres deep. The instruments send data via satellite every 18 minutes to the US Geological Survey (USGS) in Menlo Park, California, and to the University of Queensland.

Other more commonly used strain meters, called dilatometers, measure volumetric strain only – movement that occurs equally in all directions. They cannot distinguish from which direction strain is coming, or which direction it is going, unlike Gladwin’s instrument. Such measurements of the orientation of the strain are vital in California where shear strain, not volumetric strain, is dominant.

Shear strain occurs as one plate slips past the other. Gladwin’s instrument measures shear strain by detecting minute changes in the diameter of the borehole along three different axes.

Gladwin believes that what he is seeing may be part of a bigger pattern. His instrument measured a build-up of strain in the 12 months before the Loma Prieta earthquake. This build-up followed two smaller earthquakes at nearby Lake Elsman. These two earthquakes could have ‘loaded’ the fault at Loma Prieta, in much the same way as the fault near San Juan Bautista is being loaded today, he says.

But Gladwin stresses that there is no independent verification of what he is detecting because of the lack of instrumentation. However, there is cause for concern. If build-up is being measured to the south of Loma Prieta, there is every likelihood that stress is also building up farther north on the fault near San Jose, a highly populated area much closer to San Francisco.

In the San Francisco Bay Area, seismologists believe a major earthquake could occur on the Hayward Fault on the east side of the bay. This is connected to the faults near Loma Prieta by the Calaveras Fault. Two of Gladwin’s instruments, along with several volumetric meters, will be installed along the Hayward Fault as part of a project established by the USGS following the Loma Prieta earthquake. Boreholes are being drilled, and the instruments should be working by September.