Cascadia slow slip events and earthquake initiation theories: Hazards research with Plate Boundary Observatory geodetic data (Invited)
Abstract
The relationship of transient slow slip events (SSEs) to great earthquakes is a global focus of intense and critical hazards research. Plate Boundary Observatory (PBO) GPS and borehole strainmeter (BSM) networks in the Cascadia forearc provide detailed data that can be compared with simulations predicting how SSEs might evolve as a great earthquake approaches. Cascadia SSEs represent aseismic slip of a few cm in the direction of plate convergence over a period of days or weeks, in a depth range down-dip from the locked zone expected to generate the next great Cascadia subduction earthquake. During an SSE, shear stress borne in the SSE depth range is transferred up-dip at an above-background loading rate. If shear stress on the locked zone is continually accumulating, the daily probability of reaching a threshold failure stress is elevated during an SSE . Alternatively, if dynamic instability is due to rate-weakening fault strength, then SSEs still promote earthquake initiation, but that initiation may be delayed until after the SSE ends, and short-duration SSEs may have negligible effect. In some numerical simulations, great earthquakes could nucleate in the SSE depth range, where effective pressure is assumed to be low. Certain models predict that successive SSEs will slip to increasingly shallower depths, eventually encountering higher effective stress where shear heating can destabilize slip and lead to dynamic rupture. PBO GPS stations have recorded surface deformation from SSEs since inception in 2003; borehole strainmeters (BSMs) have recorded SSE strain signals since 2007. GPS and seismic tremor data show that SSEs reoccur all along the Cascadia subduction zone. An SSE is in progress somewhere in Cascadia much of the time, so the short-term probability increase warranted by a typical SSE is presumably low. We could, however, detect differences among successive SSEs and use criteria informed by the models described above to judge whether a distinctive SSE might represent a higher short-term earthquake probability increase. In all conceptual models, an SSE with more net slip and/or extending further up-dip is more likely to lead to dynamic rupture. There are also models in which faster propagation speed would promote instability by increasing the potential for shear heating. In northernmost Cascadia, BSMs near the coast, up-dip of SSEs, record transient SSE strains at high signal-to-noise ratio. Successive SSEs have differed somewhat in length and propagation speed, but not greatly in up-dip extent or net slip. BSMs up-dip of northern Oregon SSEs have recorded two large SSEs (in 2011 and 2013) having similar strain time series, as well as tremor patterns. In these regions, BSM data could allow an SSE of greater net slip, shallower up-dip extent, or unusual propagation pattern to be identified. Resolution is poorer in reaches of the forearc with BSMs only down-dip of the SSEs. Up-dip BSMs would also be best-positioned to record strain from aseismic slip approaching the locked zone. Some models predict systematic evolution of SSE behavior as a great earthquake approaches, such as decreasing intervals between SSEs, increasing rupture length and slip speed, and slip at successively shallower depths. The northern Cascadia SSEs observed with BSMs since 2007 have not exhibited these patterns, but PBO geodetic instrumentation provides an opportunity to observe them should they develop.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFM.S23C..06R
- Keywords:
-
- 1207 GEODESY AND GRAVITY Transient deformation;
- 7240 SEISMOLOGY Subduction zones;
- 4315 NATURAL HAZARDS Monitoring;
- forecasting;
- prediction;
- 1242 GEODESY AND GRAVITY Seismic cycle related deformations