17:15 〜 19:15
[SSS11-P05] Autocorrelation Functions Analysis Using Strong Motion Records from Temporary Stations in Wajima City, Ishikawa Prefecture
キーワード：The 2024 Noto Peninsula Earthquake, Wajima City, Sedimentary velocity structure, Autocorrelation functions, Strong motion observation, Strong motion records
The Noto Peninsula in Ishikawa Prefecture, located along Japan’s western shoreline, has experienced devastating structural damage due to medium to large-sized earthquakes. The Noto Peninsula earthquake (MJMA7.6, Mw 7.5) occurred on January 1, 2024, at 16:10 JST. In Wajima City, located in the north-western part of Ishikawa prefecture, more buildings were damaged by strong ground motion in the lowlands than in the hill sites. A quick comparison between the observed records at the lowland and hill sites during the mainshock found that the JMA47600 station, located in the lowland of Wajima City, experienced higher peak ground velocity (PGV) characterized by long-period motion, compared to the K-NET Wajima (ISK003) station of NIED on the hill. For instance, the PGV of the NS component at JMA47600 was approximately 1.5 times that at ISK003 despite the two stations being merely about 1km apart. The PGV of EW and UD components are also higher at the JMA47600 station. One reason that could explain the differences in waveforms and the higher PGV values at JMA47600 compared to ISK003 is the local site effects. The variation in sedimentary velocity structure is a key factor in explaining these differences in strong motion records. To investigate the differences in site amplification characteristics of sediments between the lowland and hill areas, we installed strong ground motion observation stations: five stations along the east-west line and six stations along the north-south line. The temporary observation stations were deployed and operated between June 17 and December 7, 2024. We have known the variety of the site amplification characteristics since the comparison of many aftershocks, including the largest aftershock on November 26.
By applying autocorrelation functions (ACF) to strong motion records, it is possible to estimate the depth of the layer boundary with high impedance contrast. In this study, we applied the ACF analysis and Phase Weighted Stacking (PWS) method (Schimmel and Paulssen, 1997) to clarify the ACF signal in order to determine the travel time difference between the direct and the reflected waves for the estimation of sedimentary depth. Since ACF analysis requires a high signal-to-noise ratio, we carefully selected S-wave segments while eliminating low-quality signals. We compared the observed travel time difference with the theoretical one for each station while considering the hypocentral distribution. From the PWS ACF, we identified a clear signal at our temporary station near ISK003 that appeared earlier than the one near the JMA47600. We plan to apply these procedures to all stations and examine the existing sedimentary velocity structures.
[Acknowledgments] We used strong motion records from K-NET and KiK-net, the strong motion observation networks of the NIED.
By applying autocorrelation functions (ACF) to strong motion records, it is possible to estimate the depth of the layer boundary with high impedance contrast. In this study, we applied the ACF analysis and Phase Weighted Stacking (PWS) method (Schimmel and Paulssen, 1997) to clarify the ACF signal in order to determine the travel time difference between the direct and the reflected waves for the estimation of sedimentary depth. Since ACF analysis requires a high signal-to-noise ratio, we carefully selected S-wave segments while eliminating low-quality signals. We compared the observed travel time difference with the theoretical one for each station while considering the hypocentral distribution. From the PWS ACF, we identified a clear signal at our temporary station near ISK003 that appeared earlier than the one near the JMA47600. We plan to apply these procedures to all stations and examine the existing sedimentary velocity structures.
[Acknowledgments] We used strong motion records from K-NET and KiK-net, the strong motion observation networks of the NIED.