The large-scale plate-tectonics framework of El Salvador was defined in the “plate-tectonics revolution” of the 1960s and 1970s, but important issues related to seismic hazards depend on details that have been only recently, or are not yet, understood. Present evidence suggests that coupling across the interface-thrust zone beneath coastal El Salvador is sufficient to produce occasional interface-thrust earthquakes as large as M ∼8. The rate of such earthquakes is determined by the percentage of relative plate motion that is accumulated as elastic strain on the thrust-fault interface between the Cocos and Caribbean plates, which appears to be lower than in many other subduction zones, but is not well established. Earthquakes in the interior of the Cocos plate, such as the El Salvador earthquake of January 13, 2001, account for a significant percentage of Wadati-Benioff zone earthquakes. Separate consideration of the seismic hazard posed by, respectively, Cocos intraplate earthquakes and interface-thrust earthquakes is complicated by the difficulty of separating interface-thrust and Cocos intraplate events in earthquake catalogs. Earthquakes such as the San Vicente–San Salvador sequence of February 13–25, 2001, probably result from the motion of the Central American forearc northwestward with respect to the interior of the Caribbean plate; the geometry of the fault systems that accommodate the motion remains to be worked out. Understanding of this tectonic complexity and associated seismic hazards will be facilitated greatly by the long-term operation of high-sensitivity local seismograph networks, such as that operated by, and currently being upgraded by, the Servicio Nacional de Estudios Territoriales (SNET) of El Salvador.

We present a catalog of subduction zone earthquakes along the Pacific coast from central El Salvador to eastern Chiapas, Mexico, from 1526 to 2000. We estimate that the catalog is complete since 1690 for M S ≥7.4 thrust events and M ≥ 7.4 normal-faulting events within the upper 60 km of the down-going slab. New intensity maps were constructed for the 27 earthquakes since 1690, using mostly primary data sources. By calibrating with recent events we find that the long axis of the (MM) VII intensity contour for such large earthquakes well approximates the length and location of rupture along the subduction zone and can thus be used to estimate the locations and magnitudes of older events. The section from western El Salvador to Chiapas appears to have ruptured completely in a series of four to five earthquakes during each of the periods 1902–1915, 1743–1776, and possibly 1565–1577. Earthquakes of M W 7.75 ± 0.3 have caused major damage along the 200 km long section from San Salvador to Guatemala City every 71 ± 17 yr, apparently since at least 1575. Although the January 2001 El Salvador earthquake caused damage within part of this zone, no major thrust earthquake has occurred there since at least 1915. We find that much of this section has been relatively quiescent for moderate earthquakes shallower that 50 km since at least 1963. The conditional probability that an earthquake of M W 7.75 ± 0.3 will occur at this location in the next 20 yr is estimated at 50% (±30%).

Abstract The western boundary of the Caribbean Plate is delineated by a wide band of seismicity associated with the Middle American Trench (Continental Scale Map 4, Seismicity Map of North America , this volume). Most of this seismicity occurs along the interface between the Caribbean Plate and the subducting Cocos Plate, at and beneath the Middle American Trench; this seismicity is dealt with by Dewey and Suarez (this volume). In this chapter I concentrate on seismicity within Central America that occurs both within the Caribbean Plate and along the transcurrent Caribbean-North American plate boundary (hereafter called the “CARB-NOAM boundary”); within the Caribbean Plate, seismicity occurs principally along the volcanic front (hereafter called the “volcanic zone”); other seismicity also occurs within the wedge-shaped region between the volcanic zone and the CARB-NOAM boundary, and within the complicated and poorly understood region of southern Costa Rica. Except possibly within the latter region, all such seismicity is probably confined to the upper latter 20 km of the crust (hereafter called “upper-crustal” seismicity). Unlike seismicity along the subduction zone, upper-crustal seismicity within Central America is not conspicuous in international earthquake catalogs or on global seismicity maps. This is due to the fact that, at any particular magnitude level, upper-crustal earthquakes are infrequent compared with those along the subduction zone. However, because they are very shallow and often near heavily populated areas, upper-crustal earthquakes with magnitudes as small as M 5 can produce significant damage in Central America, whereas subduction-zone earthquakes generally must have magnitudes of at least