Seismicity as a guide to global tectonics and earthquake prediction

Abstract

Studies of the precise spatial distribution of earthquakes in island arcs, along mid-ocean ridges and along transform faults have contributed significantly to recent theories and syntheses of global tectonics. This paper reviews evidence from seismicity for transform faulting on ridges, underthrusting of the lithosphere in arcs, and delineation of large lithospheric plates by the world-wide distribution of earthquakes. That earthquakes along fracture zones are restricted almost exclusively to the region between two spreading ridge segments is one of the most compelling pieces of evidence for transform faulting and sea-floor spreading. Seismic phenomena are generally explained as a result of interactions at or near the edges of large lithospheric plates. Most seismic activity in island arcs is confined to a narrow zone less than a few tens of kilometers thick that dips under the arc to depths as great as 700 km. This activity approximately defines the configuration of downgoing slabs of lithosphere. The distribution of hypocenters of earthquakes at depths greater than about 400 km indicates that the lithospheric slab is contorted and deformed at these depths in most arcs. This greater deformation may be related to the penetration of the lithosphere into a more viscous part of the upper mantle. Seismic activity in the Tonga island arc of the southwest Pacific where two oceanic plates converge is extremely simple in that nearly all of the activity is confined to a simple planar zone and little or no seismic activity is found above the main dipping zone. Moderate seismic activity is found above the dipping seismic zone in Japan, South America, Kamchatka, and New Zealand, where continental crust is being underthrust by an oceanic lithospheric slab. Some activity behind arcs may delineate a zone of extension as proposed by Karig (1970), but the rear sides of some other island arcs are characterized by compressional mechanism solutions. Several zones of moderate to high seismic activity, such as in the St; Lawrence valley in eastern North America and in the northeastern Indian Ocean, may represent intra-plate deformation with high in-situ stresses. In the second part of this report the mapping of the spatial extent of aftershock zones of large earthquakes yields valuable clues to the prediction of earthquakes in island arcs and along transform faults. The accurate mapping of aftershock locations (unlike the mere plotting of epicenters of large earthquakes) approximately defines zones of rupture. Plate tectonic theory indicates that gaps in activity for large earthquakes for the past tens to hundreds of years are likely sites of future large earthquakes. These gaps are of high priority for study and instrumentation. Large earthquakes appear to be much more regular than smaller shocks in their distribution with respect to space, time and size. Aftershock zones of large earthquakes tend to abut without significant overlap, even for rupture zones as long as 1200 km.