2:45 PM - 3:00 PM
[SSS11-22] Estimation of Heterogeneous Attenuation Structure in Kyushu Region Based on Block Inversion Analysis and Estimation of Short-Period Level of Hyuga-Nada Earthquake
Keywords:GIT, Hyuga-Nada Earthquake, Short-Period Level
1. Introduction
The Kyushu area has complex propagation path characteristics due to the propagation of seismic waves inside the Philippine Sea Plate and the influence of low attenuation areas around active volcanoes. In addition, an interplate earthquake of MJ 7.1 on August 8, 2024 and MJ 6.6 on January 13, 2025 occurred in the Hyuga-Nada, and it is considered that studying the source characteristics of these earthquakes will contribute to improving the accuracy of future earthquake motion prediction. Therefore, in this study, we estimate the heterogeneous attenuation structure in the Kyushu area in the depth direction by block inversion analysis, and also estimate the short-period level of earthquakes that occur in the Kyushu area.
2. Data and method
We collected records of trench earthquakes that occurred in the Kyushu area and were observed by K-NET and KiK-net, meeting the following conditions. The period covered was from the start of observation to January 15, 2025, the MW of F-net was greater than 4.5, the epicenter depth was less than 200 km, the epicenter distance was less than 200 km, and the ground surface records were less than 200 cm/s2 to avoid the influence of nonlinearities in the ground. The distribution of target earthquakes and observation points is shown in Figure 1. Fourier amplitude spectra for the time interval between S-wave arrival and earthquake size (8 s for MW<6.0, 20 s for MW>=6.0) were used for the analysis, and the frequency range of interest was 0.5 to 20 Hz.
Propagation path characteristics in the block inversion analysis were modeled in three layers (continental crust, mantle, and Philippine Sea plate) in the depth direction, taking into account the horizontal heterogeneity of each layer. The horizontal block size for each layer is 0.2° mesh. The propagation distance of each layer was evaluated by 2D ray tracing on the section connecting the epicenter and the observation point. S-wave velocities were set to 3.4 km/s, 4.5 km/s, and 4.7 km/s, respectively. Examples of ray traces are shown in Figure 2. From earthquake 1 to KGS010, a large fraction of the propagation is in the mantle, and from earthquake 1 to MYZ013, a large fraction is in the interior of the Philippine Sea Plate. Earthquake 2 is dominated by propagation in the continental crust, but its horizontal location is different.
The trade-off between source and site characteristics was resolved by using the theoretical amplification factor (E+F)/2E at the OITH09 and KGSH12 ground station locations as a constraint.
The optimized ground model is based on Kasamatsu et al. (2014).
3. Results
The estimated inhomogeneous damping structure is shown in Figure 3. The color in the figure indicates the Q value at 10 Hz. In the continental crust, low-Q regions were estimated around volcanic zones. In the continental crust, low-Q regions were estimated around volcanic zones. In the continental crust, low-Q areas were estimated in the southeastern part of Miyazaki Prefecture and around the volcanic zones. This trend is consistent with Komatsu and Oda (2015). The qualitative trend is similar to that of Tomozawa et al. (2021), who estimated the heterogeneous attenuation structure only for earthquakes in the inland crust. The Q-values of the mantle are generally lower than those of the other layers, and those of the Philippine Sea Plate are relatively higher than those of the other layers, with generally homogeneous Q-values.
The short-period levels of the Hyuga-nada earthquakes on August 8, 2024 and January 13, 2025 were estimated to be 2.15 × 1019 Nm/s2 and 1.22 × 1019 Nm/s2, respectively. Intra-plate earthquakes tend to have higher short-period levels on average than inter-plate earthquakes. In the future, we will conduct envelope inversion analysis for these earthquakes to estimate the region of short-period emission.
The Kyushu area has complex propagation path characteristics due to the propagation of seismic waves inside the Philippine Sea Plate and the influence of low attenuation areas around active volcanoes. In addition, an interplate earthquake of MJ 7.1 on August 8, 2024 and MJ 6.6 on January 13, 2025 occurred in the Hyuga-Nada, and it is considered that studying the source characteristics of these earthquakes will contribute to improving the accuracy of future earthquake motion prediction. Therefore, in this study, we estimate the heterogeneous attenuation structure in the Kyushu area in the depth direction by block inversion analysis, and also estimate the short-period level of earthquakes that occur in the Kyushu area.
2. Data and method
We collected records of trench earthquakes that occurred in the Kyushu area and were observed by K-NET and KiK-net, meeting the following conditions. The period covered was from the start of observation to January 15, 2025, the MW of F-net was greater than 4.5, the epicenter depth was less than 200 km, the epicenter distance was less than 200 km, and the ground surface records were less than 200 cm/s2 to avoid the influence of nonlinearities in the ground. The distribution of target earthquakes and observation points is shown in Figure 1. Fourier amplitude spectra for the time interval between S-wave arrival and earthquake size (8 s for MW<6.0, 20 s for MW>=6.0) were used for the analysis, and the frequency range of interest was 0.5 to 20 Hz.
Propagation path characteristics in the block inversion analysis were modeled in three layers (continental crust, mantle, and Philippine Sea plate) in the depth direction, taking into account the horizontal heterogeneity of each layer. The horizontal block size for each layer is 0.2° mesh. The propagation distance of each layer was evaluated by 2D ray tracing on the section connecting the epicenter and the observation point. S-wave velocities were set to 3.4 km/s, 4.5 km/s, and 4.7 km/s, respectively. Examples of ray traces are shown in Figure 2. From earthquake 1 to KGS010, a large fraction of the propagation is in the mantle, and from earthquake 1 to MYZ013, a large fraction is in the interior of the Philippine Sea Plate. Earthquake 2 is dominated by propagation in the continental crust, but its horizontal location is different.
The trade-off between source and site characteristics was resolved by using the theoretical amplification factor (E+F)/2E at the OITH09 and KGSH12 ground station locations as a constraint.
The optimized ground model is based on Kasamatsu et al. (2014).
3. Results
The estimated inhomogeneous damping structure is shown in Figure 3. The color in the figure indicates the Q value at 10 Hz. In the continental crust, low-Q regions were estimated around volcanic zones. In the continental crust, low-Q regions were estimated around volcanic zones. In the continental crust, low-Q areas were estimated in the southeastern part of Miyazaki Prefecture and around the volcanic zones. This trend is consistent with Komatsu and Oda (2015). The qualitative trend is similar to that of Tomozawa et al. (2021), who estimated the heterogeneous attenuation structure only for earthquakes in the inland crust. The Q-values of the mantle are generally lower than those of the other layers, and those of the Philippine Sea Plate are relatively higher than those of the other layers, with generally homogeneous Q-values.
The short-period levels of the Hyuga-nada earthquakes on August 8, 2024 and January 13, 2025 were estimated to be 2.15 × 1019 Nm/s2 and 1.22 × 1019 Nm/s2, respectively. Intra-plate earthquakes tend to have higher short-period levels on average than inter-plate earthquakes. In the future, we will conduct envelope inversion analysis for these earthquakes to estimate the region of short-period emission.