Insight on rupture dynamics from probabilistic fault models of Chilean large earthquakes
Abstract
A common way to characterize seismic sources is to estimate the history of fault slip from observed co-seismic displacements. This type of modeling is usually termed "kinematic" because it does not address the underlying failure mechanism on the seismic fault. These kinematic models can then be used to inform dynamic models of fault behaviors that explicitly consider the stress associated with fault slip. However, these dynamic models often rely on assumptions regarding the failure criterion, friction on the fault, etc. To get insight on rupture dynamics, a possibility is to use kinematic models as a boundary condition to compute the temporal stress evolution on the fault (e.g., Tinti et al., JGR 2005). We explore the stress evolution observed from kinematic models of large earthquakes in Chile. In particular, we investigate the 2014 Mw=8.1 Iquique earthquake and the 2015 Mw=8.3 Illapel earthquake. This study is carried out in a Bayesian framework, where several kinematic model samples describe our knowledge of the rupture process given available observations. To calculate the shear traction, we solve a boundary integral equation method (BIEM) using slip rate history derived from kinematic model samples. This is done assuming a homogenous infinite isotropic medium and a planar fault. The retrieved shear traction evolution shows slip-weakening relation, hence we estimate the critical slip weakening distance (Dc) and the breakdown work (i.e., area under the stress-slip curve up to minimum traction). Accounting for uncertainty in the model space is useful to assess our constraints on dynamic parameters, thereby helping us to understand the physical mechanisms associated with earthquake ruptures. Preliminary results for the 2014 Iquique earthquake indicate an almost linear dependence of Dc with total slip. As mentioned in previous studies, our resolution on Dc is probably very poor given the filter bandpass used for inversion (5-100 sec). Breakdown work, on the other hand, appears to be stably estimated despite the non-uniqueness of other dynamic parameters. Therefore, we focus on scaling relationships between breakdown work and kinematic parameters such as slip amplitude and rupture velocity. This analysis is then compared with typical scaling relationships expected for crack-like and pulse-like ruptures.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMS037.0007C
- Keywords:
-
- 7209 Earthquake dynamics;
- SEISMOLOGY;
- 7215 Earthquake source observations;
- SEISMOLOGY;
- 7230 Seismicity and tectonics;
- SEISMOLOGY;
- 8118 Dynamics and mechanics of faulting;
- TECTONOPHYSICS