A Stress-Boundary Condition Model for Intraplate Seismicity
Abstract
There is no current consensus on the driving mechanism(s) and spatiotemporal progression of intraplate seismicity. Are large historical and paleo intraplate earthquakes, such as the late Holocene events that struck the New Madrid Seismic Zone (NMSZ), isolated in time or are they part of a persistent seismic cycle? How can this seismicity be maintained in the absence of interseismic strain accumulation inside continental interiors?
Here we develop a model with a vertical fault, infinite along strike, with its root in the upper mantle, characterized by negligible lower crustal frictional strength and/or crustal detachment from the upper mantle. This type of strength condition can develop, for instance, as a result of long-term upwelling of mantle volatiles along the fault zone with a stress-dependent hysteretic permeability [Garagash et al., AGU 2017]. The fault is loaded by a constant far field stress, a boundary condition (B.C.) that is more appropriate than constant velocity for a stable continental interior. We show that conditions can produce a constant stress seismic slip cycle when a rate-and-state frictional rheology characterized by mid-crustal velocity-weakening and lower-crustal strengthening is used. This cycle results from stress "yo-yoing" between the frictionally stable and unstable parts of the fault: as the former slips aseismically, it is loading and eventually nucleating seismic slip in the latter, which in turn reloads the former, and so forth. The main difference between this statically-driven cycle and a kinematically-driven (constant velocity B.C.) one is in the exponential decay of the down-dip aseismic slip rate following an earthquake. This results in near zero fault slip rate and undetectable surface strain rate over the majority of the interseismic period. Our results also show a very strong sensitivity of the fault system to the loading: for instance, a 1-2% change in the far field stress leads to a factor of 100 change in the timing of the earthquake cycle. The clustering of NMSZ paleoseismicity in the Holocene could then be the result of a drastic shortening of an otherwise very long cycle timing (>100 Kyr) in response to a slight increase in fault loading by deglaciation and/or sediment erosion processes.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.S14B..03G
- Keywords:
-
- 7209 Earthquake dynamics;
- SEISMOLOGYDE: 7212 Earthquake ground motions and engineering seismology;
- SEISMOLOGYDE: 7215 Earthquake source observations;
- SEISMOLOGYDE: 7223 Earthquake interaction;
- forecasting;
- and prediction;
- SEISMOLOGY