Arrays of stations are located along the length of the fault. Subarrays 1 and 8 are located off the end of the fault. Subarrays 2 and 7 are at the two ends of the fault, subarrays 3-6 are along the fault trace. On each side of the fault the two closest stations are 15m from the fault. The hypocenter is located beneath subarray 3.
The displacement seismograms at the closest station to the fault for each subarray are shown in Figure 13.17. Static offsets are observed on the FP component at sites adjacent to the fault. The static offsets are due to the sudden elastic rebound of the fault and are sometimes referred to as fling. The static offset on one side of the fault shows one-half of the total differential motion across the fault. The sign of the static offset is opposite at the stations on the other side of the fault. The FP component static offsets are reduced in amplitude at the two sites at the end of the fault (subarrays 2 and 7). This is due to the elastic response of the medium around the fault in which "push" and "pull" quadrants result in deformation that is not parallel to the fault strike. At these two sites, for the same reason, the FN component also shows a static offset. There are negligible static offsets at sites located off of the ends of the fault.
The FN component on the other hand only has dynamic pulse-like motions with no static offset (except for the two fault end cases already discussed). The FN component has the same amplitude on each side of the fault like the FP component, but in contrast it also has the same sign. The pulse amplitude is seen to steadily grow from subarray 3 to 6 due to directivity. At subarray 7 the FN amplitude is slightly reduced, and it remains large at subarray 8 off of the end of the fault.
Because the fault is vertically dipping and the slip direction is horizontal the vertical motions are very weak in comparison to the two horizontal components. In the other simulations for different source and slip-direction geometries dynamic pulses and static offsets are observed on the vertical component.
In velocity the FN component becomes very pronounced, growing steadily in amplitude in the direction of rupture with maximum values on the order of 1 m/s. The velocity pulses have the same sign on each side of the fault. In contrast the FP (fling) velocity pulses along the fault have constant amplitude, but opposite sign on the two sides of the fault. The FP velocity pulses also have amplitude on the order of 1 m/s. The FP peak velocity close to the fault is controlled by the slip velocity on the fault. Slip velocity is proportional to stress drop and therefore for events with stress drop on the order 10 MPa the slip velocity is on the order of 1 m/s.
The magnitude of the directivity effect depends on how co-linear the strike and slip directions are. In the case of a vertical strike-slip fault the strike and rake are co-linear producing a maximum directivity effect. In the case of a reverse fault the rake is perpendicular to the strike and this produces a minimum directivity effect (except in the updip direction). The simulations for the other fault orientations clearly illustrate this point. Figure 13.17 also shows that the amount of directivity depends on distance from the epicenter to the station.
In general the simulations show that both static offsets and dynamic directivity-controlled pulses need to be considered on any of the three components to account for possible faulting variability. In fact, as the faulting style trends to dip-slip cases there is increased vertical motions and a transition of fling-controlled motions (static offsets) to the FN component, and directivity-controlled motions to the FP component. In oblique faulting cases it will be necessary to consider both dynamic pulses and static offsets in displacement on all three components.
Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail: email@example.com
© 2005, The Regents of the University of California