Potential-field studies were conducted in seismically active regions throughout the world: in California [ Jachens and Griscom, 1995], Oregon [ Blakely et al., 1995], in the U. S. mid-continent [ Hildenbrand et al., 1992; Hildenbrand and Hendricks, 1995], in eastern Canada [ Goodacre et al., 1993; Mohajer, 1993; Thomas et al., 1993], in Colombia [ Toto and Kellogg, 1992], along the Rhine graben [ Prodehl et al., 1992], in Norway [ Karpuz et al., 1991], and in the Gulf of Aqaba [ Alamri et al., 1991].
The Loma Prieta earthquake (
) struck the San Francisco Bay Area
in October 1989 killing 63 people and causing $5.9 billion in damages
to public and private property throughout the region from San Francisco
to Santa Cruz, California. Understanding this earthquake and its many
aftershocks is critical to mitigating the effects of future earthquakes
of the area. This understanding is complicated, however, because the
Loma Prieta earthquake occurred in a region of significant geologic
complexity.
Jachens and Griscom [1995] produced a comprehensive interpretation of the epicentral region of the Loma Prieta earthquake based on magnetic and gravity data and constrained by geologic and seismic information. Their modeling experiments show a San Andreas fault system that deviates significantly from a simple vertical boundary between two major plates. Aftershocks indicate that the deeper parts of the fault surface in the epicentral region dip steeply to the southwest, but potential-field data indicate that in the shallow subsurface magnetic sediments of Pliocene age are thrust northeastward beneath rocks on the northeast side of the San Andreas fault. A few kilometers south, the San Andreas is nearly vertical in the shallow subsurface, changing to a southwest dip farther south. Jachens and Griscom, [1994] concluded that most Loma Prieta aftershocks occurred along three major faults, the San Andreas, Zayante-Vergeles, and Sargent faults, and that the widely dispersed pattern of epicenters was the result of non-vertical and spatially variable surfaces on these faults.
The New Madrid seismic zone in the Mississippi embayment was the site of a M=8 earthquake in 1811 and is estimated to be capable of producing a M=6 earthquake on the order of every 100 years. Located just 100 km from Memphis, Tennessee, such an event could have devastating consequences. Most of the seismicity of the New Madrid seismic zone occurs within the Reelfoot graben, a presumed rift active during the Cambrian and now entirely concealed beneath Upper Cretaceous and younger sedimentary rocks. The graben is defined on the basis of aeromagnetic and seismic-reflection data.
A specially designed aeromagnetic survey was flown over the northern part of the New Madrid seismic zone to investigate the Reelfoot rift and related structures [ Hildenbrand et al., 1992; Hildenbrand and Hendricks, 1995]. The survey was flown just 91 m above ground along flight lines spaced 400 m apart. The aeromagnetic data, after various enhancing operations, show numerous linear features lying parallel to the New Madrid seismic zone . These magnetic anomalies apparently reflect faults, but the physical connection between faulting and magnetic sources is complicated; namely, the sources of the anomalies lie at depths of about 1 km within the relatively nonmagnetic sedimentary overburden, several kilometers above the faulted basement. Hildenbrand et al. [1992] concluded that the magnetic sources may be igneous intrusions emplaced along the faults and into the sediments. Alternatively, faulting may have promoted the growth of magnetic minerals, either authigenic pyrrhotite or magnetite altered from pyrite, in the overlying sediments. In any case, the pattern of faults determined from the aeromagnetic data should lead to new stress models for the New Madrid seismic zone .