17:15 〜 18:45
[SCG40-P04] Modeling temporal variations of physical properties at the plate boundary in the Nankai subduction zone
キーワード：孔内、S反射波、C0002、鉛直アレイ
At the C0002 site in the Nankai subduction zone, a broadband seismometer has been equipped at a depth of 900 m from the seafloor, and a short-period seismometer has also been deployed at the seafloor. Using ambient noise records at a frequency band of 0.5–2.5 Hz observed by the vertical array of the seismometers, we were able to persistently detect S-wave (S) reflections from the megasplay fault and the top of the oceanic crust (Tonegawa et al. 2023, JpGU). We investigated temporal variations of the amplitude ratio of the reflected S wave to the direct S wave from the seafloor to the borehole sensor, and found that the reflection coefficients at the megasplay fault increased during slow slip events, ~0.2. Such a change in physical property is probably caused by fluid migration between the megasplay fault and the subducting oceanic crust. However, the degree of the velocity change is not clear, and its estimation using the reflection coefficient changes is important for quantitative estimations for the amount of fluid migration. In this study, using the numerical simulation of S reflection waveforms, we estimate velocity changes at the megasplay fault.
In the numerical simulation, we conducted a three-dimensional finite-difference approach with rotated staggered grids and second-order calculation accuracies in time and space (Saenger et al. 2000). The model space (x1–x2–x3) was 20 × 20 × 16 km3 and included four layers with a grid spacing of 0.02 km, and the x3-axis was taken as vertically downward. The source location was set to (x1, x2, x3) = (10.0, 10.0, 2.0) at the seafloor of 2 km depth. A horizontal force with a Ricker wavelet at a center frequency of 1.68 Hz (maximum frequency of 2.43 Hz) was applied to the source location. The stations were assigned at the seafloor and borehole locations (2 km and 2.9 km depths). The Vp, Vs, and density of the homogeneous medium and fluid are 5.2 km s–1, 2.89 km s–1, and 2.57 g cm–3, and 1.5 km s–1, 0.0 km s–1, and 1.0 g cm–3, respectively, in which Vp is roughly referred to an averaged Vp around a high velocity body of Shiraishi et al. (2019).
We first assigned a low velocity zone of 2 km with Vs of 1.156, 1.105, and 0.867 km/s with a constant Vp of 3.8 km/s. The dip angle of the layer is 10º and the depth from the seafloor is 6 km. The case for a Vs of 1.156 km/s reproduced the reflection coefficient for the ambient condition. It was expected that the case for the lower Vs, e.g., 0.867 km/s, reproduce the reflection coefficient increase during slow slip events. However, changes of the calculated reflection coefficient were small and the observed changes of the reflection coefficients were not reproduced. Secondly, we assigned a thin low velocity layer with a thickness of 0.1 km and a Vs of 0.867 km/s, which correspond to the megasplay fault. Vp and Vs of 4.1 km/s and 1.8 km/s are assigned to the rest of 1.9 km-layer. In this case, we successfully reproduced a substantial change of the reflection coefficient. Our results suggest that fluid migration occurs during slow slip events, and its region is limited to megasplay fault.
Acknowledgement
We thank T. Kimura for deploying the geophone at the seafloor at C0002 site. This work was supported by JSPS KAKENHI Grant No. 21H05202, 21H05204 in Scientific Research on Transformative Research Areas “Science of Slow-to-Fast earthquakes”.
In the numerical simulation, we conducted a three-dimensional finite-difference approach with rotated staggered grids and second-order calculation accuracies in time and space (Saenger et al. 2000). The model space (x1–x2–x3) was 20 × 20 × 16 km3 and included four layers with a grid spacing of 0.02 km, and the x3-axis was taken as vertically downward. The source location was set to (x1, x2, x3) = (10.0, 10.0, 2.0) at the seafloor of 2 km depth. A horizontal force with a Ricker wavelet at a center frequency of 1.68 Hz (maximum frequency of 2.43 Hz) was applied to the source location. The stations were assigned at the seafloor and borehole locations (2 km and 2.9 km depths). The Vp, Vs, and density of the homogeneous medium and fluid are 5.2 km s–1, 2.89 km s–1, and 2.57 g cm–3, and 1.5 km s–1, 0.0 km s–1, and 1.0 g cm–3, respectively, in which Vp is roughly referred to an averaged Vp around a high velocity body of Shiraishi et al. (2019).
We first assigned a low velocity zone of 2 km with Vs of 1.156, 1.105, and 0.867 km/s with a constant Vp of 3.8 km/s. The dip angle of the layer is 10º and the depth from the seafloor is 6 km. The case for a Vs of 1.156 km/s reproduced the reflection coefficient for the ambient condition. It was expected that the case for the lower Vs, e.g., 0.867 km/s, reproduce the reflection coefficient increase during slow slip events. However, changes of the calculated reflection coefficient were small and the observed changes of the reflection coefficients were not reproduced. Secondly, we assigned a thin low velocity layer with a thickness of 0.1 km and a Vs of 0.867 km/s, which correspond to the megasplay fault. Vp and Vs of 4.1 km/s and 1.8 km/s are assigned to the rest of 1.9 km-layer. In this case, we successfully reproduced a substantial change of the reflection coefficient. Our results suggest that fluid migration occurs during slow slip events, and its region is limited to megasplay fault.
Acknowledgement
We thank T. Kimura for deploying the geophone at the seafloor at C0002 site. This work was supported by JSPS KAKENHI Grant No. 21H05202, 21H05204 in Scientific Research on Transformative Research Areas “Science of Slow-to-Fast earthquakes”.