A physical model of the effect of a shallow weak layer on strong ground motion for strike-slip ruptures

Abstract

AbstractWe report results of foam rubber modeling of the effect of a shallow weak layer on ground motion from strike-slip ruptures. Computer modeling of strong ground motion from strike-slip earthquakes has in some cases involved somewhat arbitrary assumptions about the nature of slip along the shallow part of the fault (e.g., fixing the slip to be zero along the upper 2 km of the fault plane). Fault-slip inversion studies indicate that the high-frequency radiation from the shallow part of strike-slip faults is typically less than that for the deeper parts of the fault. In many cases, faults (1) may be weak along the upper few kilometers of the fault zone and may not be able to maintain high levels of shear strain required for high dynamic energy release during earthquakes and (2) may have different constitutive relations for fault slip, for example, slip strengthening. The object of this article is to present results of physical modeling using a shallow weak layer, in order to support the physical basis for assuming a long rise time and a reduced high-frequency pulse for the slip on the shallow part of faults. A weak zone was modeled by inserting weak plastic layers of a few inches in width into the foam rubber model. The long-term strength of the weak layer is about an order of magnitude less than the rest of the model. The transient strength is velocity strengthening with the strength estimated to be about three times higher at slip velocities typical of dynamic slip events. It appears a 2-km-deep, weak zone along strike-slip faults could indeed reduce the high-frequency energy radiated from shallow slip and that this effect can best be represented by superimposing a small-amplitude, short rise-time pulse at the onset of a much longer rise-time slip. For the 15-cm weak zone, the average pulse amplitude is reduced by a factor of about 0.4. The reduction factor for the 20-cm case is about 0.2. For the 30-cm case, it is about 0.1. From these results, we can see that the thicker the weak layer, the more difficult it is for a short rise-time acceleration pulse to push its way through the weak layer to the surface. The velocity strengthening property of the weak layer further damps the slip motion and increases the rise time. These results support reducing the high-frequency radiation from shallower parts of strike-slip faults in modeling studies if it is known that the shallow part of the fault is weak or has not stored up a large shear stress.