Two earthquakes of the same magnitude won't necessarily shake the ground with the same oomph. What makes a big difference, it turns out, is whether the earthquake is caused by two rock masses being shoved together--or torn apart. The finding, reported  in tomorrow's issue of Science, could have reverberations for architects, who may have to revamp design standards for buildings near active faults.
Seismologists had seen hints that the amount of shaking seems to vary with the type of motion--compressional or extensional--along sloping faults, which are common near the boundaries of tectonic plates. But no one knew why, until seismologist David Oglesby and his colleagues at the University of California, Santa Barbara, decided to study what happens as the rock begins to slip.
The researchers devised a mathematical model, which revealed that as soon as any fault begins to rupture, it generates shock waves that race through the rock ahead of the rupture. When the waves reach the surface, some get reflected back into the ground. Oglesby's co-worker Ralph Archuleta says that when those reflected waves cross the fault plane, they change the pressures on the fault. For rocks that are being pulled apart, the waves also trigger a second rupture just ahead of the initial one, cutting friction even more. By the time the rupture reaches the surface, relatively little energy is left to shake the ground.
It's a different story for faults in which the rocks get squeezed together. Here, the model showed that reflected waves jack up the friction by squeezing the rocks on opposite sides of the fault even tighter. When the rupture breaks through to the surface, the amount of energy that is released is greatly increased, causing shaking roughly twice as intense as that on a fault that's pulling rock apart. "That was unexpected," Archuleta says.
The findings will need to be considered when evaluating the earthquake risk to buildings near faults, says seismologist Steve Day of San Diego State University. Although some quake records had suggested that the type of fault influenced shaking, he says, no one could be sure that those quakes were representative of all faults: "It's extremely satisfying to now have a very well-grounded physical explanation."