5:15 PM - 7:15 PM
[SCG45-P42] Dynamic conditions of the Λ-shaped multifault rupture in the 2011 Mj7.0 Fukushima Hamadori, Japan, earthquake
Keywords:Dynamic rupture simulation, Active Fault, Regional stress field, Dynamic triggering
The 2011 Mj7.0 Fukushima Hamadori, Japan, earthquake (Hamadori earthquake) is thought to have occurred through a complex process including Λ-shaped multifault rupture. So far, several attempts have been made to investigate the spatiotemporal distribution of slip on multiple faults during the Hamadori earthquake by slip inversion using static ground displacements captured by SAR observations (e.g., Kobayashi et al., 2012) and strong motion waveforms (e.g., Hikima, 2012). On the other hand, the physical mechanism of the occurrence of multifault rupture has not been fully discussed, considering the effects of dynamic stress change and initial stress state. In this study, we perform dynamic rupture simulations of the Hamadori earthquake to elucidate the physical conditions that reproduce the observed Λ-shaped multifault rupture.
The three main physical elements that govern fault rupture processes are the friction law, the fault geometry, and the regional stress field, which need to be modeled for dynamic rupture simulations. Linear slip weakening friction law (Ida, 1972) is assumed as the friction law. Since it is difficult to constrain and verify the spatial distribution of frictional parameters by observations, each frictional parameter is given assuming a uniform value throughout faults. For the geometry of each fault (Shionohira, Itozawa, and Yunotake fault), the fault trace identified by Kobayashi et al. (2012) is used to constrain the top and the earthquake catalog constructed by Kato et al. (2013) is used to constrain the subsurface geometry, respectively. The regional stress field in the source region of the Hamadori earthquake is constrained by the crustal stress map for inland areas of Japan constructed by Uchide et al.(2022). The fault geometries and the regional stress field determine the initial stress state on the faults. The fault geometry model, the model of initial stress state, and the assumption of frictional parameters involve uncertainties. Therefore, it is necessary to conduct a parameter study to evaluate the effect of these uncertainties on the calculated rupture process.
Parameter studies on initial stress state, frictional parameters, and fault geometry revealed a parameter set that can reproduce several observed features of Hamadori earthquake. In that case, the rupture on Shionohira fault triggered the rupture on Yunotake fault due to dynamic stress change caused by the rupture on Shionohira fault. The large and transient increase in coulomb stress caused by dynamic slip of Shionohira fault in the region where Shionohira fault and Yunotake fault are closest to each other is suggested to be an important factor in triggering. The results of the parameter study on the initial stress state suggest that the bifurcation between two very different rupture scenarios, one with and the other without triggering on Yunotake fault, is controlled by the initial stress state on Yunotake fault. The parameter study on frictional parameters confirmed that triggering on Yunotake fault can occur without fine tuning to the only specific set of frictional parameters. In addition, to investigate the effect of the geometry of Yunotake fault on the rupture process, calculations were performed for both cases of assuming a curved and a planar geometry of Yunotake fault, and the results of each calculation were compared with those of the inverse analysis of observed data. Ruptures on Yunotake fault can be triggered for both cases of assuming a curved geometry and a planar geometry. On the other hand, the characteristics of the slip distribution at depth were different. From this difference, it is suggested that the depth dependence of the slip distribution on Yunotake fault may be a manifestation of the listric fault geometry.
The three main physical elements that govern fault rupture processes are the friction law, the fault geometry, and the regional stress field, which need to be modeled for dynamic rupture simulations. Linear slip weakening friction law (Ida, 1972) is assumed as the friction law. Since it is difficult to constrain and verify the spatial distribution of frictional parameters by observations, each frictional parameter is given assuming a uniform value throughout faults. For the geometry of each fault (Shionohira, Itozawa, and Yunotake fault), the fault trace identified by Kobayashi et al. (2012) is used to constrain the top and the earthquake catalog constructed by Kato et al. (2013) is used to constrain the subsurface geometry, respectively. The regional stress field in the source region of the Hamadori earthquake is constrained by the crustal stress map for inland areas of Japan constructed by Uchide et al.(2022). The fault geometries and the regional stress field determine the initial stress state on the faults. The fault geometry model, the model of initial stress state, and the assumption of frictional parameters involve uncertainties. Therefore, it is necessary to conduct a parameter study to evaluate the effect of these uncertainties on the calculated rupture process.
Parameter studies on initial stress state, frictional parameters, and fault geometry revealed a parameter set that can reproduce several observed features of Hamadori earthquake. In that case, the rupture on Shionohira fault triggered the rupture on Yunotake fault due to dynamic stress change caused by the rupture on Shionohira fault. The large and transient increase in coulomb stress caused by dynamic slip of Shionohira fault in the region where Shionohira fault and Yunotake fault are closest to each other is suggested to be an important factor in triggering. The results of the parameter study on the initial stress state suggest that the bifurcation between two very different rupture scenarios, one with and the other without triggering on Yunotake fault, is controlled by the initial stress state on Yunotake fault. The parameter study on frictional parameters confirmed that triggering on Yunotake fault can occur without fine tuning to the only specific set of frictional parameters. In addition, to investigate the effect of the geometry of Yunotake fault on the rupture process, calculations were performed for both cases of assuming a curved and a planar geometry of Yunotake fault, and the results of each calculation were compared with those of the inverse analysis of observed data. Ruptures on Yunotake fault can be triggered for both cases of assuming a curved geometry and a planar geometry. On the other hand, the characteristics of the slip distribution at depth were different. From this difference, it is suggested that the depth dependence of the slip distribution on Yunotake fault may be a manifestation of the listric fault geometry.