Intraplate Seismicity and Lithospheric Strength as Inferred from 3D Seismic Models

Abstract:

Focal mechanism studies and other stress indicators indicate that intraplate earthquakes in central and eastern North America are consistent with an ENE-WSW compressive stress field that acts on existing faults. Here we argue that either locally thinned lithosphere (as beneath the Reelfoot Rift, central US) or regional lithospheric thinning (as beneath the edge of the North American craton or the rifted continental margin) constitutes zones of relatively low lithospheric strength where crustal strain accumulates. We use seismic surface waves to determine the 3D shear-wave seismic velocity structure of the lithosphere, and find that the Reelfoot Rift is underlain by a zone with low mantle seismic velocities that extends to at least 200 km depth. Thus, the Reelfoot Rift, which hosts the New Madrid Seismic Zone, is unique among North American paleo-rifts in term of the properties in the mantle. We hypothesize that this low-velocity mantle volume is weaker than its surroundings and that the Reelfoot Rift consequently has relatively lower elastic plate thickness that would tend to concentrate tectonic stress within this zone. On a continental scale, the 3D velocity model clearly identifies an approximately 220-km-thick, high seismic velocity lithospheric root beneath the North America craton which has a low rate of crustal seismicity and very few events with Mw≥6. We attribute the relatively aseismic nature of the craton to dry, cold conditions within the cratonic lower crust and mantle lithosphere. Conversely, we find that a high proportion of intraplate events are concentrated around the pronounced lateral gradient in lithospheric thickness that surrounds the craton. We attribute this observation to a lateral decrease in lithospheric strength at the edge of the North American craton. This relationship between intraplate seismicity and lithospheric properties is apparent in maps that compare regional and continental lithospheric thickness with crustal seismicity. We conclude that we should not expect intraplate earthquakes to be uniformly distributed. They are found to be associated with zones of reduced lithospheric strength that can be identified in 3D seismic models. This conclusion lends support to the concept that historical seismicity is a useful indicator of potential future earthquake locations.