A probabilistic framework for hazard assessment and mitigation of induced seismicity related to deep geothermal systems
Abstract
Slip on tectonic faults take place over a wide range of spatial and temporal scales as earthquakes, continuous aseismic creep, or transient creep events. Shallow creep events on continental strike-slip faults can occur spontaneously, or are coupled with earthquake afterslip, or are triggered by nearby earthquakes. Despite more than five decades of observations, the mechanism of shallow creep events and their implications for seismic hazard are still not fully understood. To understand the mechanism of creep events, we developed a physics-based numerical model to simulate shallow creep events on a strike-slip fault with rate-and-state frictional properties (Wei et al., 2013). We show that the widely used synoptic model (Scholz, 1998) cannot reproduce both rapid afterslip and frequent creep events as observed on the Superstition Hills fault in the Salton Trough after the 1987 Mw 6.6 earthquake. Rather, an unstable layer embedded in the shallow stable zone is required to match the geodetic observations of the creep behavior. Using the strike-slip fault model, we studied the triggering process of creep events, by either static or dynamic, or combined stress perturbations induced on the fault by nearby earthquakes. Preliminary results show that static stress perturbations in the effective normal stress on a system with spontaneous creep events can advance or delay creep events. The magnitude and timing of perturbations determines the clock change of creep events. The magnitude and interval of creep events changes permanently after static stress perturbation. Dynamic stress perturbations in effective normal stress can advance the timings of creep events when the perturbation temporally decreases the effective normal stress. A threshold exists for instantaneous triggering. The size of triggered slip increases as the dynamic perturbation increases in the direction of less normal stress. The system returns to pre-perturbation state after a long period of no slip. The length of the recovery time depends on the size of triggered slip therefore the magnitude and duration of perturbation. Perturbations that temporally increase effective normal stress do not have significant influence on the timings of future creep events. We applied our theoretical models to the Salton Trough, California, where both shallow creep events and earthquakes are common. We systematically analyzed the level of dynamic and static triggering from nearby earthquakes for the last 30 years, including moderate (> M5) to large (>M6) earthquakes. By incorporating these triggering to our fault model, we are trying to understand 1) which mechanism is dominant, static or dynamic; 2) whether a critical threshold exists, like in the generic model with synthetic dynamic perturbations for the instantaneous triggering of shallow creep events in Salton Trough; 3) the effect of fault orientation with respect to the incoming seismic waves. By developing state-of-the-art models and constraining parameters with rich datasets from Southern California, we aim to transition from a conceptual understanding of fault creep towards a quantitative and predictive understanding of the physical mechanism of creep events on continental strike-slip faults.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFM.S44B..06W
- Keywords:
-
- 7223 SEISMOLOGY / Earthquake interaction;
- forecasting;
- and prediction;
- 7230 SEISMOLOGY / Seismicity and tectonics;
- 7230 SEISMOLOGY Seismicity and tectonics;
- 7250 SEISMOLOGY Transform faults;
- 4316 NATURAL HAZARDS Physical modeling;
- 1207 GEODESY AND GRAVITY Transient deformation