5:15 PM - 6:45 PM
[U15-P30] Three-dimensional Attenuation Structure in the Noto Peninsula, Japan
Keywords:the 2024 Noto Peninsula earthquake, S-wave attenuation structure, tomography, strong ground motion
Introduction
We examine the 3-D S-wave attenuation (Qs) structure using short-period strong motion records in central Japan, including the epicenter area of the 2024 Noto Peninsula Earthquake (Mj7.6) (hereafter, the M7.6 Noto earthquake).
The northern part of the Noto Peninsula has historically experienced relatively large earthquakes, such as the 1729 Noto-Sado earthquake (M 6.6-7.0) and the 2007 Noto Peninsula earthquake (Mj 6.9). There has also been earthquake swarm activity in the eastern Noto Peninsula since December 2020, with an Mj 6.5 earthquake occurring in May 2023. Nishimura et al. (2023) suggested that prolonged aseismic slip due to upwelling fluid triggered intense earthquake swarms. The presence of fluid beneath the seismic swarm region was discussed as a low S-wave velocity and high-Vp/Vs anomalies (Nakajima, 2022).
The Qs value is a sensitive parameter for the presence of fluid. Indeed, distinct low-Qs regions have been detected directly beneath island arc volcanoes in Japan (e.g., Nakamura and Uetake, 2004; Nakamura 2009). Estimating Qs has the analytical advantage of a smaller trade-off between structure and source location, in contrast to seismic wave velocity tomography. Thus, the 3-D Qs structure is valuable for investigating the fluid distribution around the Noto Peninsula.
Data/Method
We used tomographic analysis with spectral inversion (e.g., Nakamura & Uetake, 2002; Nakamura, 2009). The study area is 35°-39° N and 136°-140° E. The modeling space was divided into blocks with size of 0.1°× 0.1°×10 km. The data used in the tomography analysis were S-wave spectrum amplitudes calculated from the acceleration records obtained by K-NET and KiK-net stations at the free surface of the ground.
We collected the earthquakes that meet following the criteria: (1) earthquakes with magnitudes of Mj 7.5 or less that occurred from May 1996 to January 2024, (2) earthquakes with focal mechanisms determined by F-net, and (3) earthquakes with focal depths shallower than 200 km. Then, we collected seismic waveforms recorded at stations within 100 km and 500 km of the epicentral distances for the earthquakes occurred shallower and greater than 30 km, respectively to avoid the influence of Lg waves[MI1] . To eliminate the influence of nonlinear responses, we excluded seismic waveforms with acceleration amplitudes exceeding 100 cm/s^2. Based on criteria (1), the Mj 7.6 Noto earthquake was excluded from the analysis. Finally, we obtained 47,455 S-wave spectral amplitudes from 885 earthquakes.
Results and discussion
We carried out checkerboard resolution test (CRT) to evaluate the spatial resolution of our tomography analysis. The CRTs provided a resolution[MI2] of 20 km in the horizontal direction and 10 km in the vertical direction. However, since further study is required on the accuracy of the Qs value itself, we will focus on the qualitative characteristics of the obtained Qs structure.
The obtained Qs structure revealed two significant low-Qs regions in the northeastern part of the Noto Peninsula at a depth of 0-20 km, near the epicenters of the M7.6 Noto earthquake and the 2007 Noto earthquake. The locations of the low-Qs regions correspond well with those of the low-Vs region (Nakajima, 2022). Additionally, it was found that the zonal high-Qs region parallel to the low-Qs in the southern part of the Noto Peninsula.
In our inversion, we estimated the source terms and site amplification factors along with the Qs structure simultaneously. It is important to note that there may be a trade-off depending on the distributions of seismic stations and earthquakes. For instance, if the seismic source terms are overestimated, significant low-Qs value would be derived around the source area. Therefore, in this presentation, we will verify the reliability of the estimated 3-D Qs structure in more detail and examine the relationship between the crustal attenuation structures and seismic activity in the Noto Peninsula.
We examine the 3-D S-wave attenuation (Qs) structure using short-period strong motion records in central Japan, including the epicenter area of the 2024 Noto Peninsula Earthquake (Mj7.6) (hereafter, the M7.6 Noto earthquake).
The northern part of the Noto Peninsula has historically experienced relatively large earthquakes, such as the 1729 Noto-Sado earthquake (M 6.6-7.0) and the 2007 Noto Peninsula earthquake (Mj 6.9). There has also been earthquake swarm activity in the eastern Noto Peninsula since December 2020, with an Mj 6.5 earthquake occurring in May 2023. Nishimura et al. (2023) suggested that prolonged aseismic slip due to upwelling fluid triggered intense earthquake swarms. The presence of fluid beneath the seismic swarm region was discussed as a low S-wave velocity and high-Vp/Vs anomalies (Nakajima, 2022).
The Qs value is a sensitive parameter for the presence of fluid. Indeed, distinct low-Qs regions have been detected directly beneath island arc volcanoes in Japan (e.g., Nakamura and Uetake, 2004; Nakamura 2009). Estimating Qs has the analytical advantage of a smaller trade-off between structure and source location, in contrast to seismic wave velocity tomography. Thus, the 3-D Qs structure is valuable for investigating the fluid distribution around the Noto Peninsula.
Data/Method
We used tomographic analysis with spectral inversion (e.g., Nakamura & Uetake, 2002; Nakamura, 2009). The study area is 35°-39° N and 136°-140° E. The modeling space was divided into blocks with size of 0.1°× 0.1°×10 km. The data used in the tomography analysis were S-wave spectrum amplitudes calculated from the acceleration records obtained by K-NET and KiK-net stations at the free surface of the ground.
We collected the earthquakes that meet following the criteria: (1) earthquakes with magnitudes of Mj 7.5 or less that occurred from May 1996 to January 2024, (2) earthquakes with focal mechanisms determined by F-net, and (3) earthquakes with focal depths shallower than 200 km. Then, we collected seismic waveforms recorded at stations within 100 km and 500 km of the epicentral distances for the earthquakes occurred shallower and greater than 30 km, respectively to avoid the influence of Lg waves[MI1] . To eliminate the influence of nonlinear responses, we excluded seismic waveforms with acceleration amplitudes exceeding 100 cm/s^2. Based on criteria (1), the Mj 7.6 Noto earthquake was excluded from the analysis. Finally, we obtained 47,455 S-wave spectral amplitudes from 885 earthquakes.
Results and discussion
We carried out checkerboard resolution test (CRT) to evaluate the spatial resolution of our tomography analysis. The CRTs provided a resolution[MI2] of 20 km in the horizontal direction and 10 km in the vertical direction. However, since further study is required on the accuracy of the Qs value itself, we will focus on the qualitative characteristics of the obtained Qs structure.
The obtained Qs structure revealed two significant low-Qs regions in the northeastern part of the Noto Peninsula at a depth of 0-20 km, near the epicenters of the M7.6 Noto earthquake and the 2007 Noto earthquake. The locations of the low-Qs regions correspond well with those of the low-Vs region (Nakajima, 2022). Additionally, it was found that the zonal high-Qs region parallel to the low-Qs in the southern part of the Noto Peninsula.
In our inversion, we estimated the source terms and site amplification factors along with the Qs structure simultaneously. It is important to note that there may be a trade-off depending on the distributions of seismic stations and earthquakes. For instance, if the seismic source terms are overestimated, significant low-Qs value would be derived around the source area. Therefore, in this presentation, we will verify the reliability of the estimated 3-D Qs structure in more detail and examine the relationship between the crustal attenuation structures and seismic activity in the Noto Peninsula.