Rupture zones of large South American earthquakes and some predictions

Abstract

This study attempts to forecast likely locations for large shallow South American earthquakes in the near future by examining the past space-time pattern of occurrence of large (M ≥ 7.7) earthquakes, the lateral extent of their rupture zones, and, where possible, the direction of rupture propagation. Rupture zones of large shallow earthquakes generally abut and do not overlap. Patterns of rupture propagation appear to follow certain trends. These facts, plus the nonrandom behavior of the space-time history of seismic activity, present consistencies that may permit prediction, in a gross sense, of future events. By mapping the rupture zones of large earthquakes (in contrast with plotting only epicenters), it is possible to identify segments of the shallow seismic zone that have not ruptured in many decades. Limited experience elsewhere indicates that these gaps between rupture zones tend to be filled by large-magnitude earthquakes. In certain places it is possible to make approximate estimates of the time of occurrence of the next large earthquake. For at least 300 or 400 years, the entire fault segment near the Central Valley province of central and southern Chile (about 32°–46°S) has fractured about once each century from a generally N-S progression of several large (M ≥ 8) earthquakes. Large earthquakes in this region have almost always occurred to the south of a previous large earthquake. In addition, it is possible to infer a direction of rupturing for two large earthquakes in this century (1928 and 1960). Both these earthquakes fractured southward away from the rupture zone of an earlier earthquake. It would be consistent with these observations if a new series started about the end of this century near Valparaiso (33°S) and progressed southward. In other sections of South America there are several extensive segments of the active seismic belt that have not ruptured during this century. Northern Chile and southernmost Peru (about 17°–25°S) have been relatively aseismic for about 100 years. South of Lima (about 12.5°–14°S), between the rupture zones of the 1940 and 1942 Peruvian earthquakes, there is another significant gap in recent activity. Both these regions are probably areas of relatively high earthquake risk. The northern Peru and southern Ecuador region (about 9°–1°S) has also been relatively aseismic during this century. However, this region differs from the two previously mentioned gaps in that this coastal zone was a region of moderate seismicity during historic times. Perhaps aseismic creep is an unusually important factor in relieving tectonic strain along this particular segment of the shallow seismic zone. Another possibility is that large shallow earthquakes in this region have an extremely long recurrence time. Much of the shallow seismic zone of northern Ecuador and southwestern Colombia has ruptured twice during this century. During large earthquakes in this region, the rupturing tends to be directed toward the north or NE. The data for this region suggest that the area to the NE of the 1958 Colombian earthquake may be a region of relatively high earthquake risk.