1:45 PM - 3:15 PM
[SSS06-P14] Investigation of multi-segment earthquake on the Median Tectonic Line active fault zone based on dynamic rupture simulation (Part 3)
Keywords:the Median Tectonic Line active fault zone, dynamic rupture simulation, multi-segment earthquake
The Median Tectonic Line fault zone (MTL) is one of the most active fault zones in Japan. To investigate the possibility and conditions that a multi-segment earthquake occur at MTL in Shikoku, southwest Japan, we construct a dynamic source model based on geological and geophysical data and perform dynamic rupture simulations. In the previous study (Kase and Urata, 2022), a large spatial variation of shear stress is derived from the stress field model that reflects the heterogeneous orientation of the principal stress axis along the fault strike and thus, not only a multi-segment rupture but also an single-segment rupture is unlikely to occur. In this study, we introduce a heterogeneous distribution of friction coefficients into the stress model, considering the faulting history of MTL, and investigate the possibility of a multi-segment earthquake.
The fault and medium models are the same as those used in the previous study. The current stress field model reflects the heterogeneous orientation of the principal stress axes along the strike as reported in the previous study. The depth profile of the stress drop is modeled as follows: the stress drop and strength excess are proportional to depth up to the depth of 10 km, the strength excess is constant below 10 km, and the stress drop decreases with depth to zero at the depth of 15 km and negative below 15 km. We then obtain the depth profile of the stress drop that satisfies the scaling law between fault length and slip (Matsuda et al., 1980) from preliminary calculations.
The ratio of the average slip rate of each segment (HERP, 2017) is roughly consistent with the ratio of slip according to the scaling law (Matsuda et al., 1980), and thus we assume that the average stress drop of each segment is the same value. Although the average event interval of each segment has a double-half variation, the median interval is about 1100 years, and the latest events occurred at the generally same period, before around the 16th century which is indicated by the historical record (HERP, 2017). Therefore, assuming that the standard event interval of each segment is 1100 years and the latest events occurred 440 years ago, and considering that the shear stress just after the rupture is equal to the dynamic frictional stress, we can assume that the shear stress just before the next earthquake is the current shear stress plus 60% of the average stress drop. Similarly, the dynamic frictional stress can be assumed to be the current shear stress minus 40% of the average stress drop, and the dynamic frictional coefficient is calculated by ignoring the time variation of the normal stress. The difference between the static and dynamic friction coefficients is assumed to be a constant (Kawakata, personal communication).
For the boundary conditions of the fault plane, Coulomb's rupture criterion and the friction constitutive law of slip weakening (Ida, 1972; Andrews, 1976) are applied, and the rupture processes on the fault plane are calculated by the difference method (Kase and Day, 2006).
The homogenous stress drop distribution in the strike direction by introducing a heterogeneous distribution of friction coefficients and the constant difference between the static and dynamic friction coefficients result in a heterogeneous ratio of strength excess to stress drop not only in the strike but also in the dip direction. Thus, heterogeneity of strength excess dominates simulation results. The appropriate difference between the static and dynamic friction coefficients should be investigated, and a sensitivity of the standard event interval and the time elapsed since the most recent event should also be considered for each segment.
Acknowledgments: This research was funded by Research Project for Long-term Evaluation Methods of Multi-segment Earthquakes from Active Fault Zones in FY2022 by MEXT.
The fault and medium models are the same as those used in the previous study. The current stress field model reflects the heterogeneous orientation of the principal stress axes along the strike as reported in the previous study. The depth profile of the stress drop is modeled as follows: the stress drop and strength excess are proportional to depth up to the depth of 10 km, the strength excess is constant below 10 km, and the stress drop decreases with depth to zero at the depth of 15 km and negative below 15 km. We then obtain the depth profile of the stress drop that satisfies the scaling law between fault length and slip (Matsuda et al., 1980) from preliminary calculations.
The ratio of the average slip rate of each segment (HERP, 2017) is roughly consistent with the ratio of slip according to the scaling law (Matsuda et al., 1980), and thus we assume that the average stress drop of each segment is the same value. Although the average event interval of each segment has a double-half variation, the median interval is about 1100 years, and the latest events occurred at the generally same period, before around the 16th century which is indicated by the historical record (HERP, 2017). Therefore, assuming that the standard event interval of each segment is 1100 years and the latest events occurred 440 years ago, and considering that the shear stress just after the rupture is equal to the dynamic frictional stress, we can assume that the shear stress just before the next earthquake is the current shear stress plus 60% of the average stress drop. Similarly, the dynamic frictional stress can be assumed to be the current shear stress minus 40% of the average stress drop, and the dynamic frictional coefficient is calculated by ignoring the time variation of the normal stress. The difference between the static and dynamic friction coefficients is assumed to be a constant (Kawakata, personal communication).
For the boundary conditions of the fault plane, Coulomb's rupture criterion and the friction constitutive law of slip weakening (Ida, 1972; Andrews, 1976) are applied, and the rupture processes on the fault plane are calculated by the difference method (Kase and Day, 2006).
The homogenous stress drop distribution in the strike direction by introducing a heterogeneous distribution of friction coefficients and the constant difference between the static and dynamic friction coefficients result in a heterogeneous ratio of strength excess to stress drop not only in the strike but also in the dip direction. Thus, heterogeneity of strength excess dominates simulation results. The appropriate difference between the static and dynamic friction coefficients should be investigated, and a sensitivity of the standard event interval and the time elapsed since the most recent event should also be considered for each segment.
Acknowledgments: This research was funded by Research Project for Long-term Evaluation Methods of Multi-segment Earthquakes from Active Fault Zones in FY2022 by MEXT.