5:15 PM - 7:15 PM
[SSS11-P04] Estimation of Subsurface Structure in Noto Peninsula by Love Wave Microtremor Array Method
Keywords:Microtremor array survey, Love wave
In this study, we conducted microtremor array observations at the Togi Branch Office in Shika Town, where significant ground shaking was recorded during the 2024 Noto Peninsula Earthquake. Subsurface structure was estimated from the phase velocities of both Rayleigh and Love waves. We use Love wave phase velocity obtained from horizontal motion, as an additional constraint to construct a more reasonable subsurface structure model.
The observations were conducted on September 13 and November 19, 2024, using seven microtremor sensors (Hakusan JU410) at the Togi Branch Office in Shika Town, Ishikawa Prefecture. A dual equilateral triangular array was deployed with radii of 4 m (S-array), 30 m (M-array), and 150 m (L-array), with sampling frequencies of 100 Hz or 200 Hz. Due to the significant non-stationary noise on the L-array data, only the S-array and M-array data were used for analysis.
The phase velocity of Rayleigh waves was estimated from vertical components using the SPAC method. For Love waves, rotational components of the horizontal motion were calculated, and the SPAC method was applied for the rotational components to derive phase velocities. The observed data were divided into segments of 2048 samples, shifted by 1024 samples, with 20 segments grouped into one block, and a total of five blocks analyzed. The phase velocities and standard deviations were calculated.
The estimated phase velocities of the Rayleigh waves were 230–490 m/s in the 2.8–20 Hz range. The ones of the Love waves were 300–280 m/s at 3.1–5 Hz and approximately 240 m/s at 12–15 Hz. The maximum standard deviation was 23 m/s at 2.9 Hz for Rayleigh waves and 22 m/s at 3.1 Hz for Love waves. Love wave phase velocities could be estimated within the 3–6 Hz and 13–16 Hz ranges but were not obtained in other frequency bands. Beyond 4 Hz, standard deviations decreased, resulting in more stable phase velocity estimations.
For subsurface structure estimation, we made the deep subsurface structure model from the National Research Institute for Earth Science and Disaster Resilience (J-SHIS) as a base model. The shallowest layer (Vs = 650 m/s) was divided into three sublayers, and their parameters were adjusted with observed phase velocity. We found a well-fit model with a first-layer S-wave velocity of 350 m/s and a thickness of 20 m, and a second-layer S-wave velocity of 350 m/s with trial-and-error.
To evaluate the validity of the estimated subsurface structure model, the root mean square errors (RMSEs) between the observed and theoretical phase velocities of Rayleigh and Love waves were calculated. By using the phase velocities of the Love wave, the range of possible models was more limited. The optimal S-wave velocity structure has a first-layer thickness ranging from 18 to 36 m and a second-layer thickness ranging from 1 to 40 m. This study demonstrates that adding Love wave constraints can improve the accuracy of subsurface structure estimation.
Acknowledgement: We thank Profs. Kimiyuki Asano (Kyoto University), Susumu Kurahashi (Aichi Institute of Technology) and Mr. Kazuhiro Somei (Geo-Research Institute) for helping us with observation. This study was supported by the Core-to-Core Collaborative research program of the Earthquake Research Institute, The University of Tokyo and the Disaster Prevention Research Institute, Kyoto University.
The observations were conducted on September 13 and November 19, 2024, using seven microtremor sensors (Hakusan JU410) at the Togi Branch Office in Shika Town, Ishikawa Prefecture. A dual equilateral triangular array was deployed with radii of 4 m (S-array), 30 m (M-array), and 150 m (L-array), with sampling frequencies of 100 Hz or 200 Hz. Due to the significant non-stationary noise on the L-array data, only the S-array and M-array data were used for analysis.
The phase velocity of Rayleigh waves was estimated from vertical components using the SPAC method. For Love waves, rotational components of the horizontal motion were calculated, and the SPAC method was applied for the rotational components to derive phase velocities. The observed data were divided into segments of 2048 samples, shifted by 1024 samples, with 20 segments grouped into one block, and a total of five blocks analyzed. The phase velocities and standard deviations were calculated.
The estimated phase velocities of the Rayleigh waves were 230–490 m/s in the 2.8–20 Hz range. The ones of the Love waves were 300–280 m/s at 3.1–5 Hz and approximately 240 m/s at 12–15 Hz. The maximum standard deviation was 23 m/s at 2.9 Hz for Rayleigh waves and 22 m/s at 3.1 Hz for Love waves. Love wave phase velocities could be estimated within the 3–6 Hz and 13–16 Hz ranges but were not obtained in other frequency bands. Beyond 4 Hz, standard deviations decreased, resulting in more stable phase velocity estimations.
For subsurface structure estimation, we made the deep subsurface structure model from the National Research Institute for Earth Science and Disaster Resilience (J-SHIS) as a base model. The shallowest layer (Vs = 650 m/s) was divided into three sublayers, and their parameters were adjusted with observed phase velocity. We found a well-fit model with a first-layer S-wave velocity of 350 m/s and a thickness of 20 m, and a second-layer S-wave velocity of 350 m/s with trial-and-error.
To evaluate the validity of the estimated subsurface structure model, the root mean square errors (RMSEs) between the observed and theoretical phase velocities of Rayleigh and Love waves were calculated. By using the phase velocities of the Love wave, the range of possible models was more limited. The optimal S-wave velocity structure has a first-layer thickness ranging from 18 to 36 m and a second-layer thickness ranging from 1 to 40 m. This study demonstrates that adding Love wave constraints can improve the accuracy of subsurface structure estimation.
Acknowledgement: We thank Profs. Kimiyuki Asano (Kyoto University), Susumu Kurahashi (Aichi Institute of Technology) and Mr. Kazuhiro Somei (Geo-Research Institute) for helping us with observation. This study was supported by the Core-to-Core Collaborative research program of the Earthquake Research Institute, The University of Tokyo and the Disaster Prevention Research Institute, Kyoto University.