KGS Home > Geologic Hazards
Ground-Motion Simulation

Estimation of ground motion in the central United States is not straightforward because of a scarcity of moderate to large earthquakes. Therefore, synthetic strong-ground motion simulation is the only way to provide additional information about the potential damage moderate or large earthquakes could cause. The most widely used simulation methods to generate synthetic ground motion for the central United States are the stochastic point and finite source simulations of high-frequency ground motions. Because the stochastic source model lacks a physical process, such as source rupture process, directivity effect, and wave propagation, there is a certain limitation in the ground-motion modeling. In addition, the challenges of performance-based engineering will increasingly require that structures be modeled as complex, dynamic, nonlinear multi-degree-of-freedom systems. This requires the entire time histories of strong ground motion with three input components.

In several recent investigations, a kinematic model combined with empirical or theoretical Green's function computational techniques has successfully predicted ground motion with a realistic appearance of waveform and frequency content. This composite source model has three important advantages over the pure stochastic source model. First, the slip pulse distributed on the fault inherits a stochastic property and obeys a given fractal law. Thus, the complex earthquake physical process can be partly simulated. Second, the wave propagation effect is considered by computing the theoretical Green's function based on the elasto-dynamics equation for a layered solid structure. Third, a multi-component ground motion can be generated through a computation of wave propagation in a given layered crustal structure.

For strong-motion simulation, the composite source model is described with superposition of circular subevents, which are randomly distributed on the main fault. Therefore, the subevents are allowed to overlap with each other, and the total area of the subevents is much larger than the area of the main fault. As a result, multiple triggering of subevents is generally used in order to achieve seismic moment conservation. The multiple triggering involved in composite source modeling is the same technique used in the stochastic finite source model in order to keep moment conservation. We have enhanced the composite source model and simplified the problem with squared subevent distributions. The number of subevents with characteristic dimension greater than R is proportional to R-2 . The subevents do not overlap each other, and the sum of their areas equals the area of the main fault. Each subevent is allowed to slip only once, at the arrival time of the rupture front. For the near-fault strong ground motion, the composite source model simulation can characterize the near-source directivity effect and S-wave radiation pattern properly, consistent with the theoretical consideration.