5:15 PM - 6:45 PM
[SSS03-P09] Spatio-temporal seismic structural variations and its interpretation revealed from seismic interferometry using OBS records around the northern Japan Trench
Keywords:Seismic Interferometry, Ocean bottom seismometer, Northern Japan Trench subdcuction zone, Spatio-temporal variation in seismic structure
We have applied single-station seismic interferometry analysis to the continuous records of ocean bottom seismometer (OBS) networks at off-Erimo (2006–2007) and off-Sanriku (2007–2008) along the northern Japan Trench (Sato et al., SSJ, 2023). We previously detected coseismic velocity drops of ~0.3% and subsequent increases related to M6-class local earthquakes. We interpreted that these seismic velocity changes are caused by ground damage due to strong ground shaking and its recovery. In contrast, no seismic velocity changes were detected during the tectonic tremor activities (Takahashi, 2021; Kawakubo, 2021) beneath the networks. Our result does not agree with the observation in the Hikurangi subduction zone, where seismic velocity changes were observed in association with slow earthquake activities (Wang et al., 2022). Two reasons were considered for not detecting changes associated with slow earthquakes in the northern Japan Trench: First is that the magnitude of the slow activities is smaller. And the second is that the hypocentral region is deeper than those in the Hikurangi subduction zone. In this presentation, we discuss the recovery process after seismic velocity drop and the temporal variation of correlation coefficients.
To quantify the time constant of the recovery process after the M6-class earthquakes, we fit a model function, which describes the exponential recovery process, to the time series of seismic velocity using a grid search approach. As a result, while we confirmed time constants are larger than 0.03–0.1 years at most stations, we could not constrain the longer side of the constant except for at the station in the off-Sanriku network at which the constant is estimated to be 0.07 years (error range: 0.03–0.5 years). The estimated value is slightly smaller than those observed after the 2014 off-Mie earthquake (0.27–0.45 years, Ikeda and Tsuji, 2018) and the 2008 Iwate-Miyagi Nairiku earthquake (0.1–2.6 years, Hobiger et al., 2014). Because the analyzed period in this study is 6 months and thus is considerably shorter than those in the previous studies (1 year 7 months, 3 years), our results may reflect the recovery process on a shorter time scale immediately after the earthquakes (Hobiger et al., 2012; Illien et al., 2023).
To quantify the time constant of the recovery process after the M6-class earthquakes, we fit a model function, which describes the exponential recovery process, to the time series of seismic velocity using a grid search approach. As a result, while we confirmed time constants are larger than 0.03–0.1 years at most stations, we could not constrain the longer side of the constant except for at the station in the off-Sanriku network at which the constant is estimated to be 0.07 years (error range: 0.03–0.5 years). The estimated value is slightly smaller than those observed after the 2014 off-Mie earthquake (0.27–0.45 years, Ikeda and Tsuji, 2018) and the 2008 Iwate-Miyagi Nairiku earthquake (0.1–2.6 years, Hobiger et al., 2014). Because the analyzed period in this study is 6 months and thus is considerably shorter than those in the previous studies (1 year 7 months, 3 years), our results may reflect the recovery process on a shorter time scale immediately after the earthquakes (Hobiger et al., 2012; Illien et al., 2023).