3:45 PM - 4:00 PM
[HDS10-08] Analysis of Tsunami Generation Due to Seafloor Subsidence Around the Sofu Seamount
Keywords:tsunami, Torishima, caldera, 3D calculation
The source of multiple earthquakes that occurred near Torishima on October 9, 2023, is estimated to be around the Sofu Seamount. Seafloor topographic surveys conducted in 1987 and 2023 revealed not only the formation of a caldera-like structure several kilometers in scale but also localized subsidence reaching several hundred meters horizontally and up to 400 meters vertically (Fujiwara et al., 2024). Kubota et al. (2024) and Takemura et al. (2024) focused on the large-scale caldera structure and suggested that its uplift may have generated tsunamis. However, localized subsidence is also considered a potential tsunami generation mechanism.
In this study, we conducted numerical simulations using a three-dimensional model to elucidate the tsunami generation mechanism caused by localized seafloor subsidence around the Sofu Seamount. The three-dimensional simulations employed the single-phase Navier-Stokes equation model (CADMAS-SURF/3D), developed by Arikawa et al. (2005), which calculates the ocean surface using the volume of fluid (VOF) method. Specifically, we applied a vertically downward velocity at the subsidence point by varying the subsidence rate according to a sine function over a half-cycle and analyzed the resulting tsunami wave height and period. As a calculation condition, the time-integrated value of the applied velocity at the subsidence point was set to match the observed topographic changes.
For tsunami propagation from the caldera, we coupled this model with the nonlinear long-wave equations under the hydrostatic assumption (STOC-ML), developed by Tomita et al. (2016). We conducted multiple cases by varying the duration of subsidence completion and analyzed simulation results for subsidence durations ranging from 4 to 10 minutes in 1-minute increments. The results showed that shorter subsidence durations led to higher wave heights, while significant changes in the wave period were not observed across different cases. This is considered to be due to the effects of the bay geometry.
A comparison between the observed waveforms at Hachijojima and the simulation results indicated that assuming a subsidence duration of approximately 2 to 5 minutes yielded the highest consistency with the observed data. Meanwhile, multiple seismic waves were detected in the observations, suggesting the possibility that subsidence occurred in multiple stages (Sandanbata et al., 2024; Takemura et al., 2024). However, this study assumes a single occurrence of subsidence, and the effects of multiple subsidence events remain a subject for future research.
This method enables a detailed assessment of the impact of seafloor subsidence due to submarine volcanic activity on tsunami generation, contributing to the understanding of tsunami generation mechanisms. Additionally, it has potential applications for improving future tsunami warning systems. The Sofu Seamount region has a history of seafloor topographic changes due to volcanic activity, posing a continuous tsunami generation risk. Strengthening monitoring efforts while considering the geological background is essential. This study provides valuable insights for improving the accuracy of disaster prediction and developing effective countermeasures against oceanic disasters caused by volcanic activity.
In this study, we conducted numerical simulations using a three-dimensional model to elucidate the tsunami generation mechanism caused by localized seafloor subsidence around the Sofu Seamount. The three-dimensional simulations employed the single-phase Navier-Stokes equation model (CADMAS-SURF/3D), developed by Arikawa et al. (2005), which calculates the ocean surface using the volume of fluid (VOF) method. Specifically, we applied a vertically downward velocity at the subsidence point by varying the subsidence rate according to a sine function over a half-cycle and analyzed the resulting tsunami wave height and period. As a calculation condition, the time-integrated value of the applied velocity at the subsidence point was set to match the observed topographic changes.
For tsunami propagation from the caldera, we coupled this model with the nonlinear long-wave equations under the hydrostatic assumption (STOC-ML), developed by Tomita et al. (2016). We conducted multiple cases by varying the duration of subsidence completion and analyzed simulation results for subsidence durations ranging from 4 to 10 minutes in 1-minute increments. The results showed that shorter subsidence durations led to higher wave heights, while significant changes in the wave period were not observed across different cases. This is considered to be due to the effects of the bay geometry.
A comparison between the observed waveforms at Hachijojima and the simulation results indicated that assuming a subsidence duration of approximately 2 to 5 minutes yielded the highest consistency with the observed data. Meanwhile, multiple seismic waves were detected in the observations, suggesting the possibility that subsidence occurred in multiple stages (Sandanbata et al., 2024; Takemura et al., 2024). However, this study assumes a single occurrence of subsidence, and the effects of multiple subsidence events remain a subject for future research.
This method enables a detailed assessment of the impact of seafloor subsidence due to submarine volcanic activity on tsunami generation, contributing to the understanding of tsunami generation mechanisms. Additionally, it has potential applications for improving future tsunami warning systems. The Sofu Seamount region has a history of seafloor topographic changes due to volcanic activity, posing a continuous tsunami generation risk. Strengthening monitoring efforts while considering the geological background is essential. This study provides valuable insights for improving the accuracy of disaster prediction and developing effective countermeasures against oceanic disasters caused by volcanic activity.