Geoscience Reference
In-Depth Information
Recent Advances—Dynamic Fault Modeling
Significant breakthroughs are also being made in our understanding of seismic
radiation from earthquakes. Some of the most exciting quantifications of global
earthquake ruptures in the past decade have come from innovative use of regional
arrays, showing the expansion of fault rupture for the recent immense earthquakes in
Sumatra, Chile, and Japan (e.g., Ishii et al., 2005; Lay et al., 2010a). Global seismic
network data are now used to estimate slip distribution for all major faults; geodetic
data sets using GPS, InSAR, uplift, and tsunami excitation have improved constraints
on fault displacements for events around the world. Faults likely to rupture at super
shear velocities have been identified (Bouchon and Vallee, 2003), and the complexity
of faulting beyond simple slip pulse models has been resolved (e.g., Lay et al.,
2010b). However, many fundamental questions about earthquakes remain:
•
How do earthquakes initiate?
•
What controls the branching of rupture or the triggering of one fault by
rupture of another?
•
Why does a rupture stop?
•
When and why do rupture speeds exceed the seismic shear velocity?
•
Can rupture attributes be anticipated based on geodetic determinations of prior
fault locking and strain accumulation?
To mitigate risk from earthquakes, it is first necessary to know how strongly
the ground will vibrate. This is difficult to predict given both the complexity of
earthquake ruptures and the wave focusing and defocusing effects and soil
interactions of seismic waves. At present, ground motions during earthquakes are
usually characterized by very simple measurements, such as peak ground acceleration
or velocity. These data are used by engineers to estimate the strength of ground
shaking expected during an earthquake of given size by using empirical relationships
based on past earthquake data in a given region (size of earthquake, distance to the
rupture, and local geology). This approach seems to work adequately for moderate
earthquakes, but rupture finiteness and wave directionality effects for large events
greatly complicate the ground motion prediction. Because large (and very large)
earthquakes occur infrequently, the empirical-based seismic hazard relationships are
not well constrained, and recent earthquakes have offered repeated surprises in terms
of the intensity of ground shaking actually experienced. It is desirable to move
forward from empirical approaches to quantitative modeling approaches.
Most seismic and geodetic models of fault slip are kinematic in nature;
simplifying assumptions are made to allow the estimation of the relevant parameters
(e.g. faults are planar, slip is unidirectional). Physical properties of the fault are
typically not modeled because of their complexity. However, new simulations have
shown the potential to bridge the gap from standard kinematic models to physics-
based models (e.g., Dunham and Archuleta, 2005). Dynamic rupture modeling
considers the joint stress-slip evolution during earthquake shear failure as being
driven by the redistribution of stored strain energy and can serve as the foundation for
predicting both fault behavior and strong ground motion. Dynamic rupture modeling
Search WWH ::
Custom Search