Over the past decade, the U.S. Geological Survey (USGS) has been investigating the cause of the Charleston, S.C., earthquake of 1886 and the likelihood of future earthquakes of similar magnitude (m_b 6.9–7.2, Bollinger, 1977). As part of that work, multichannel reflection surveys were started in 1979 in the Charleston area, on 1 and (Behrendt and others, 1981; Hamilton and others, 1983) and offshore (Behrendt and others, 1983). The data for lines across the continental margin were tied into the USGS offshore seismic grid in the area discussed by Dillon and others (1979). At about the same time (1978–79), Consortium for Continental Reflection Profiling (COCORP) lines in Georgia and in the Charleston, S. C., area were recorded (Cook and others, 1979; Cook and others, 1981; Schilt and others, 1983). The COCORP data for Georgia (Cook and others, 1979) and other reflection data to the northeast, as discussed by Harris and Bayer (1979), indicated the presence of the Appalachian décollement, extending seaward from the Appalachian Mountains. The authors of these papers inferred that the Appalachian décollement might extend across the Piedmont and Coastal Plain to the continental shelf. Subsequently, Iverson and Smithson (1982) suggested, on the basis of their reprocessing of the COCORP line in Georgia, that the décollement was rooted in the area of the Kings Mountain and the Carolina slate belts.

The multichannel seismic-reflection data for the Charleston, S. C., area (Behrendt and others, 1981, 1983; Schilt and others, 1983) provided evidence, particularly strong offshore, of the existence of a reflecting surface at a depth of 11.4±1.5 km that was suggested as a décollement. Behrendt and others (1981, 1983) suggested that the Charleston earthquake of 1886 might have been caused by movement on the décollement or on associated listric faults. Seeber and Armbruster (1981) suggested that movement on the Appalachian décollement, if it continued coastward to Charleston, might have caused the Charleston earthquake of 1886. The best determined focal depths for recent seismicity, from data recorded by a seismograph network operated by the USGS in the Charleston, S. C., area since 1973, are shallower than 13+2 km (Tarr and others, 1981; and Tarr and Rhea, 1983), or above the suggested décollement.

The seismic-reflection data have also shown the existence of several Triassic (?) basins beneath the Coastal Plain Late Cretaceous and Cenozoic sedimentary rock section (Behrendt and others, 1981; Behrendt, 1983; Costain and Glover, 1983; Hamilton and others, 1983; Petersen and others, 1984). The basins, in several cases, appear to be bounded by high-angle normal faults. Some of these faults may have been reactivated in Late Cretaceous and Cenozoic time as apparently reverse faults. Also they are suggested to be listric onto the décollement, thereby bearing a causal relation to Charleston seismicity (Behrendt, 1983; Behrendt and others, 1983).

The question of whether the Appalachian décollement is continuous to the coast is, therefore, important not only for the general understanding of the tectonics of the southeastern United States but for an understanding of the earthquake-hazard question as well. For these reasons, three long, deep-crustal, multichannel seismic-reflection profiles (S4, S6, and S8) were obtained by the USGS to address the problem. This report presents illustrations of interpretations of the profiles discussed by Behrendt (in press).

The Appalachian décollement does not appear continuous from the Appalachian Mountains to the coast but rather appears to extend southeastward only to the Carolina slate belt. A series of reflections on lines S4, S6, and S8 and on the COCORP line is interpreted as evidence of southeastward-dipping imbricate faults, from the Brevard fault on the northwest to beyond the Augusta fault, which marks the southeastern extent of the Eastern Piedmont fault zone. The Carolina slate belt is characterized on the four seismic profiles by a complex series of diffractions and reflections extending from less than 1 s to 8 s. These arrivals are possibly the result of layering in the metasedimentary rocks complexly disrupted by the imbricate faults. A number of Triassic (?) basins are apparent in the reflection data for the rifted Charleston terrane identified from low-gradient magnetic anomalies.