This paper describes and interprets a newly-recognized 40-km-long seismogenic zone, which is inferred to have been the locus of a damaging earthquake in 1897. That shock was the second largest known to have occurred in the southeastern United States (MMI VIII, m^b estimated at 5.8, felt over 725,000 km₂). It struck Giles County in southwestern Virginia, and a recurrence would affect populous regions on and near the central Atlantic seaboard. This paper attempts to aid in evaluating that hazard by presenting and synthesizing new seismological data with geological inferences and deductions.

A five-station, 60-km aperture seismic network has been in operation in the Giles County locale since early 1978. For the subsequent 3-year monitoring period, 10 microearthquakes (M < 2) have been detected. Eight of those 10 events, plus an additional 4 relocated felt earthquakes (3.2 < M < 4.1; 1959-1976) have a tabular distribution centered at Pearisburg, Virginia. That distribution is about 40 km long, 10 km wide, strikes N. 43° E., and has a nearly vertical extent of from 5 km to 25 km in depth. Thus, a Giles County seismogenic zone is defined presently by 12 earthquakes that span 4 orders of magnitude (0 < M < 4) and 2 decades of time (1959-1980). We conclude that the 1897 earthquake occurred on that seismogenic zone. From the orientation of the tabular zone, from evidence that greatest horizontal compressive stress trends east-northeasterly at seismogenic depths in and near Giles County and from sparse P-wave first-motion data, we infer that the monitored microseismicity probably occurs by right-reverse motion on the seismogenic zone, with the southeast side dropping down with respect to the northwest side.

In the Giles County locale, the upper 3-6 km of the crust are Paleozoic sedimentary rocks that have moved some tens of kilometers northwest on nearly horizontal detachment faults. The above-mentioned reliable hypocenters for the region lie below the deepest likely detachment, indicating that Giles County seismicity probably has no simple relationship to surface geology.

Since Precambrian time, three deformational episodes could have formed steep faults under today's surface structures, at the observed hypocentral depths. These episodes were as follows: (1) As the Iapetus Ocean (Atlantic's predecessor) opened in late Precambrian or early Paleozoic time, northeast-striking normal faults formed, probably at the inferred Iapetan continental edge in central Virginia and at least as far northwest of that locus as Giles County. (2) In late Paleozoic time, detachment faults loaded the crust with several kilometers of overthrust sedimentary rocks, perhaps forming northeast-striking thrust-load faults in a brittle analogue of isostatic depression caused by thrust masses and much lighter continental glaciers. (3) As the Atlantic Ocean opened in Mesozoic time, other northeast-striking normal faults formed on the present continental margin and inland of it.

The N. 43° E.-striking seismogenic zone seems most likely to have resulted from compressional reactivation of an Iapetan normal fault, which also may have been reactivated by late Paleozoic compression and Mesozoic extension. First, the seismogenic zone probably does not occur on a thrust-load fault. The zone underlies detached structures of southern Appalachian orientations (east-northeast), but those structures are not known to be displaced where they cross the zone. Thus, if the zone occurs on a thrust-load fault, the fault and its coeval causative central Appalachian detachments would pre-date the southern Appalachian structures. That deduction contradicts stratigraphic and structural estimates of relative ages of southern and central Appalachian detachments. Second, the zone probably does not result from a Mesozoic normal fault, because known locations of Mesozoic normal faults and grabens are well to the southeast of Giles County.

Not yet known is where else in the East reactivated Iapetan normal faults might generate shocks similar to that of 1897. However, our analysis enables us to suggest specific geological and geophysical investigations that may produce results useful in answering that question. Such investigations can concentrate on defining the area of probable occurrence of other Iapetan normal faults, and on determining whether the one inferred to underlie Giles County is uniquely active or is typical of others that might exist elsewhere.