Next: Numerical Example: Friendswood Crosswell Up: Basics of Traveltime Tomography Previous: SIRT Method

Numerical Example: Seismic CAT Scan of an Ancient Earthquake by 3-D Refraction Tomography

Earthquakes along the Wasatch Mountains constitute a grave threat to the lives of over 2 million people who live along the Wasatch Front in Utah. The typical event originates from a normal fault earthquake, where the downthrown side of a valley (see upper RHS of Figure 4) drops down relative to the upward moving mountains. It is extremely important to determine the size and recurrence intervals of large (>M6.5) earthquakes to mitigate the earthquake hazard.

Earthquake geologists try to determine the size and recurrence intervals of ancient earthquakes by digging trenches across earthquake scarps, the ancient scars of earthquakes. Figure 4 shows a scarp recently left by a large earthquake in Idaho, and a trench excavated across a fault in the Phillipines.

The earthquake geologist examines the geologic cross-section for signs of past faulting activity and correlates this activity with, typically, radiocarbon dates of soil in the trench can give an estimate of the recurrence interval of an ancient large earthquake. Such information is invaluable in estimating the earthquake hazard of a region and the time frame for a future large earthquake.

For normal faults, such as those that occur in the Intermountain Seismic Belt, trench excavations are used to identify the presence and shape of colluvial wedges. A colluvial wedge is a wedge-like zone that is filled with rubble immediately following a surface rupturing event, and is the characteristic geologic signature of an ancient earthquake that ruptured the ground surface. An example of a colluvial wedge in Mongolia is given in Figure 4, where one side of the scarp contains a thick section of rubble not seen on the other side. Dotted white line outlines colluvial wedge, and arrows indicate direction of block reverse fault movement.

Figure 5 depicts the sequence of geologic cross-sections prior to and following a normal-fault earthquake. Here, larger earthquakes produce greater displacement along the fault, so that wedge thickness is proportional to earthquake magnitude. In addition, the depth interval between neighboring wedges is proportional to the recurrence interval between large events, assuming a constant sedimentation rate. This assumes a constant sedimentation rate. Thus a typical trench study across the, e.g. Wasatch Fault, will attempt to delineate the locations and thicknesses of colluvial wedges and use this information to determine earthquake recurrence intervals and magnitudes.

The problems with trench excavations are that they are expensive, environmentally intrusive so that they are limited to sparsely populated zones, and are typically limited to depths of about 9 m. This limited depth range restricts the study of ancient earthquakes to a limited span of geologic time. Moreover, a trench only reveals a 2-D cross-section of the geologic record and so it cannot fully assess the geologic effect of the oblique-slip movement.

To overcome the limitations of the trenching methods, geophysicists have attempted to use seismic methods to image the shallow structure of faults. A new imaging method for paleoseismology is the seismic tomography method. The key idea is that the first arrival traveltimes are not affected by near-surface scattering or surface wave problems, and so could be used to image the colluvial wedge. The implicit assumption is that the seismic velocity of the colluvial wedge was significantly slower than the surrounding alluvium, as shown in Figure 6.

To perform a seismic tomography experiment we selected the Oquirrh Fault site shown in Figure 7. Here the geophones and sources were located on a 150'x35' grid of 300 points. Each shot was recorded by 300 geophones, and the first arrival times were picked and inverted to give the tomogram shown in Figure 8.

Note the remarkable resemblance between the trench log and the tomogram. Figure 9 depcits the comparison between the reflection record and the trench log.

Next: Numerical Example: Friendswood Crosswell Up: Basics of Traveltime Tomography Previous: SIRT Method
Gerard Schuster
1998-07-29