California Institute of Technology

Finite Fault Ruptures Algorithm

Large earthquakes (M>7.0), with rupture lengths of tens to hundreds of kilometers, cause damaging ground shaking in much larger areas than moderate events. Despite their rare occurrence, many people and users could benefit from an EEW system during large earthquakes (Heaton, 1985).

We are developing algorithms to automatically detect and map finite fault ruptures in real-time by performing rapid near- and far-source classifications of ground motion amplitudes in dense seismic networks (Yamada et al., 2007), and by applying image recognition techniques with sets of pre-calculated templates for variable rupture length and orientation (Böse et al., subm.) The knowledge on source dimensions will make predicted shaking intensities more accurate and thus more useful for EEW. We have developed methods that allow us simulating the acceleration envelope functions of ground motions for large earthquakes (Yamada and Heaton, 2008), which help us verifying our models.


Figure 1. Near- and far-source classified stations along with the predicted 2-D rupture for the 1999 M7.6 ChiChi earthquake in Taiwan at four time steps after rupture nucleation (Böse et al., subm.).

We found from simulated rupture series generated from stochastic models of spatially heterogeneous slip, that the recognition that rupture is occurring on a spatially smooth fault has a stronger effect than the observation of a large slip amplitude (Böse and Heaton, 2010). An EEW system for large earthquakes will thus benefit from an automated fault association (Böse et al., subm.). If rupture is occurring along a smooth fault, such as the San Andreas Fault, a warning should be issued immediately, because the probability for a large earthquake is high.

The rupture process of large earthquakes can be imaged by back-projection of high-frequency ground motions recorded by dense seismic arrays (Meng et al., 2011; Simons et al., 2011; Meng et al, 2012). The principle is analogous to the location and tracking of moving sources by antennas in a variety of military and civilian applications. We are developing the concept of a highly clustered seismic network made of multiple seismic antennas, each composed of a dozen low-cost accelerometers with a footprint of less than 1 km. This system would track in real-time the rapidly moving position of an earthquake rupture front to provide an estimate of rupture size as the earthquake unfolds.

While current EEW systems rely on seismic instrumentation, the inclusion of geodetic data from a real-time Global Positioning System (GPS) with high rate (1 hertz or higher) and low latency (seconds or less) will play a vital role in EEW for large earthquakes in the future. The integration of real-time GPS measurements into EEW systems is an ongoing effort.

People

Thomas Heaton, Jean-Paul Ampuero, Maren Böse, Egill Hauksson, Asaf Inbal, Lingsen Meng

Publications