17:15 〜 18:45
[SSS08-P16] 豊後水道におけるスロースリップ震源域直下のスラブ内間隙流体圧の時間変化
キーワード：長期的スロースリップ、間隙流体圧
At the plate boundaries beneath the Bungo Channel in southwest Japan, aseismic slip called the long-term slow slip event (SSE) has occurred repeatedly with the recurrence intervals of several years. The occurrence of SSEs is often related to over-pressurized fluids dehydrated in the subducting Philippine Sea slab. In this study, we aimed to investigate the relationship between pore-fluid pressures in the slab and the occurrence of SSEs, by estimating pore-fluid pressures in the slab beneath the source region of the SSEs.
We firstly selected focal mechanism data of earthquakes that occurred in the slab during the periods from Jan. 1997 to May 2023, listed in the F-net Seismic Moment Tensor Catalogue of National Research Institute for Earth Science and Disaster Resilience. Then, applying the data to the method of CMT data inversion (Terakawa and Matsu’ura, 2008), we estimated the stress orientations in the Philippine Sea slab in southwest Japan. Then, we estimated pore-fluid pressures which triggered the events in the slab, examining the focal mechanism data relative to the stress orientations in the 3-D Mohr diagram (Terakawa et al., 2010). For this process, we calculated angles between observed slip vectors and directions of shear stresses on the two nodal planes and determined the true fault plane with the smaller misfit angle. We evaluated the degree of over-pressures with the overpressure coefficient C (Terakawa and Hauksson, 2018), which is defined as the excess pore-fluid pressures over hydrostatic normalized by the difference between lithostatic and hydrostatic. When pore-fluid pressures are hydrostatic and lithostatic, the values of Cs are 0 and 1, respectively.
In order to estimate the pore-fluid pressure in the slab beneath the Bungo channel, we took a cross-section across the source regions of SSEs (e.g., Takagi et al., 2019) and selected 195 focal mechanism data within 50 km from the cross-section. Based on the aforementioned method together with these data, we estimated pore-fluid pressures that triggered the events and examined temporal changes in the overpressure coefficients Cs. We found that the values of Cs tended to be larger in the periods with no occurrence of SSEs (the inter-SSE periods) than those in the periods with SSEs (the intra-SSE periods). The average values of Cs were 0.1423± 0.0141 and 0.1130 ± 0.0217 in the inter-SSE periods and in the intra-SSE periods, respectively. This indicates that the pore-fluid pressures in the slab enhanced in the inter-SSE periods and decreased in the intra-SSE periods.
We also examined the temporal changes in pore-fluid pressures beneath the slab in the western Shikoku and the Hyuga-nada regions. The average values of Cs in the whole study period were 0.1356 ± 0.0120, 0.1125 ± 0.0139 and 0.1871 ± 0.0165 in the Bungo channel, the western Shikoku, and the Hyuga-nada regions, respectively. These results indicate that the average pore-fluid pressures in the slab have a positive correlation with the total slip by the SSEs. The increase in pore-fluid pressures triggers events by decreasing the fault strength. Active seismicity enhances the permeability in the slab. In the more permeable slab, fluids can be more easily transported upward to the source region of SSEs at the plate interfaces. The more abundant fluids may cause SSEs more frequently.
We firstly selected focal mechanism data of earthquakes that occurred in the slab during the periods from Jan. 1997 to May 2023, listed in the F-net Seismic Moment Tensor Catalogue of National Research Institute for Earth Science and Disaster Resilience. Then, applying the data to the method of CMT data inversion (Terakawa and Matsu’ura, 2008), we estimated the stress orientations in the Philippine Sea slab in southwest Japan. Then, we estimated pore-fluid pressures which triggered the events in the slab, examining the focal mechanism data relative to the stress orientations in the 3-D Mohr diagram (Terakawa et al., 2010). For this process, we calculated angles between observed slip vectors and directions of shear stresses on the two nodal planes and determined the true fault plane with the smaller misfit angle. We evaluated the degree of over-pressures with the overpressure coefficient C (Terakawa and Hauksson, 2018), which is defined as the excess pore-fluid pressures over hydrostatic normalized by the difference between lithostatic and hydrostatic. When pore-fluid pressures are hydrostatic and lithostatic, the values of Cs are 0 and 1, respectively.
In order to estimate the pore-fluid pressure in the slab beneath the Bungo channel, we took a cross-section across the source regions of SSEs (e.g., Takagi et al., 2019) and selected 195 focal mechanism data within 50 km from the cross-section. Based on the aforementioned method together with these data, we estimated pore-fluid pressures that triggered the events and examined temporal changes in the overpressure coefficients Cs. We found that the values of Cs tended to be larger in the periods with no occurrence of SSEs (the inter-SSE periods) than those in the periods with SSEs (the intra-SSE periods). The average values of Cs were 0.1423± 0.0141 and 0.1130 ± 0.0217 in the inter-SSE periods and in the intra-SSE periods, respectively. This indicates that the pore-fluid pressures in the slab enhanced in the inter-SSE periods and decreased in the intra-SSE periods.
We also examined the temporal changes in pore-fluid pressures beneath the slab in the western Shikoku and the Hyuga-nada regions. The average values of Cs in the whole study period were 0.1356 ± 0.0120, 0.1125 ± 0.0139 and 0.1871 ± 0.0165 in the Bungo channel, the western Shikoku, and the Hyuga-nada regions, respectively. These results indicate that the average pore-fluid pressures in the slab have a positive correlation with the total slip by the SSEs. The increase in pore-fluid pressures triggers events by decreasing the fault strength. Active seismicity enhances the permeability in the slab. In the more permeable slab, fluids can be more easily transported upward to the source region of SSEs at the plate interfaces. The more abundant fluids may cause SSEs more frequently.