Dynamic modeling of the 1999 Chi-Chi (Mw7.6) earthquake: constraints from derived constitutive relationship of kinematic modeling, and stress condition from fault zone drilling
Abstract
For two decades, since the 1999 Chi-Chi earthquake, the modeling of earthquake kinematics had been well developed. But nowadays, some questions related to earthquake dynamics still remains unknown. In view of the comprehensive study of the 1999 Chi-Chi earthquake since 1999, we tried to investigate the dynamics rupture behavior of this event to understand what controls the slip patterns as with large slip to the north and smaller amount of slip to the south. Whether the energy partition derived from dynamic modeling could be a good indicator or not, while comparing with the fracture energy directly derived from grain size distribution of the major slip zone from retrieved fault zone gouge of Taiwan Chelunpu-fault Drilling Project (TCDP). In this study, we construct a 3D dynamic rupture model using finite element method. We first obtained the constitutive relationship of the subfaults from the temporal and spatial slip distribution of the kinematic models of Ji et al. (2003). We thus estimated the dynamic parameters (e.g., apparent slip-weakening distance,) and to determine the state of stress (e.g., initial normal stress, ) distribution on the fault to perform the dynamic modeling for earthquake dynamic rupture. We found the direct adoption of the dynamic parameters from derived constitutive relationships failed in proceeding the earthquake rupture. We, thus, designed a series of numerical experiments on homogeneous and heterogeneous model, for the stress condition derived from geophysical logging analysis of TCDP, which assume that the fault ruptures with spatially uniform or non-uniform frictional behavior. After various models with different set of dynamic parameters, we find that the parameters for our best-fit heterogeneous model are =0.5~1.7(m), =1.9~7.0(m) scaled down by =0.25, S=0.3~7.5,=10~92(MPa) and =3.6~32.3(MPa). The best-fit model can simulate a rupture similar to the kinematic study and the maximum slip (~13 m) occurs in the northern part of fault. The results suggest that could be overestimated due to influence of other dynamic process related to heat or pore pressure for large earthquakes, and the different value of in the northern and southern part of fault might be a key to control the slip pattern.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.S41B..06M
- Keywords:
-
- 5199 General or miscellaneous;
- PHYSICAL PROPERTIES OF ROCKS;
- 7209 Earthquake dynamics;
- SEISMOLOGY;
- 7215 Earthquake source observations;
- SEISMOLOGY;
- 7290 Computational seismology;
- SEISMOLOGY