[S-SS07] Seismic wave propagation: Theory and Application

[SSS07-07] Formularization of SPAC between DAS and Seismometer and influence of Love waves to Rayleigh wave analysis by seismic interferometry with DAS

★Invited Papers

*Shun Fukushima^1,2, Masanao Shinohara², Kiwamu Nishida², Akiko Takeo², Tomoaki Yamada², Kiyoshi Yomogida³ (1.University of Tokyo,Graduate school of Science, Department of Earth and Planetary Science, 2.Earthquake Research Institute,The University of Tokyo, 3.Hokkaido University,Graduate school of Science,Department of Natural History Sciences)

Keywords:Distributed acoustic sensing, Seismic interferometry, Seafloor cable

Recently, Distributed Acoustic Sensing (DAS), which enables long-range observations over tens of kilometers at intervals of several meters using optical fibers as sensors, is being applied to earth sciences (e.g., Zhan 2019). We have applied seismic interferometry to DAS data in submarine cables to estimate the S-wave velocity structure of the sedimentary layer and uppermost crust with high resolution using the extracted surface waves by seismic interferometry (Fukushima et al. 2022). Previous studies have assumed that the extracted surface waves consist of Rayleigh waves only. However, Nakahara et al., (2021) showed both Rayleigh waves and Love waves are extracted when seismic interferometry is performed on strain records. The extracted Love wave has a large amplitude with a small observational distance compared to the wavelength. Although contaminations of Love waves should affect the accuracy of the phase velocity, it is difficult to separate Rayleigh wave and the Love wave from extracted surface waves by interferometry. Therefore, the influence of Love wave for phase-velocity estimation cannot be estimated because DAS has only one horizontal-component strain record.
In this study, we formulated the SPAC between DAS and 3-component seismometers observed ground motions and showed the Rayleigh wave and the Love wave can be separated by combining a three-component seismometer and DAS. Furthermore, we demonstrated that Rayleigh waves were exclusively extracted using the seismic interferometry between DAS data and a vertical displacement of seismometers. In other words, we could measure precise phase velocities of Rayleigh waves without the influence of the Love wave. We also indicated that the Love wave mainly can be derived by the seismic interferometry between DAS data and a transverse record of seismometers.
We also applied seismic interferometry to DAS data and a vertical record of ocean bottom seismometers (OBSs) connected Sanriku seafloor cable which was also used for DAS measurement. Finally, we evaluate the effect of Love wave on the phase velocity estimation using the seismic interferometry of DAS. We compared results of phase velocity estimation in the same section using two methods, (i) phase velocities were estimated from seismic interferometry between DAS records and DAS records (DAS-DAS) and (ii) Phase velocities estimated from the interferometry between DAS records and veridical component of seismometer (DAS-Vertical). Errors of phase velocities were estimated by the bootstrap method. The estimated phase velocities between 0.14 and 0.20 Hz from the two methods coincided within the margin of the errors. We confirmed that influence of Love waves is neglected for phase velocity estimation of Rayleigh waves even using a seismic interferometry of DAS-DAS records in this frequency range. However, the estimated phase velocities using DAS-DAS data are faster than the phase velocities using the method of DAS-Vertical at the frequency range from 0.10 to 0.14 Hz. According to formularization, it is concluded that the larger phase velocities result from the influence of Love waves. The developed method in this study not only enables accurate estimations of Rayleigh waves phase velocity without the influence of Love waves but also opens a new opportunity to estimate reliable three-dimensional S-wave velocity structure.