10:15 〜 10:30
[SSS04-06] Shear wave velocity structure in deep Ontong Java Plateau region by a joint finite-frequency inversion of body and surface wave data
★Invited Papers
Seismic tomography has become an essential tool for mapping the Earth's interior, providing constraints on temperature, composition, mineralogy, tectonic history, and dynamic behavior. Conventional body wave travel-time inversion based on ray theory can be considered an infinite-frequency approach. Another powerful method currently in use, full-waveform inversion, utilizes a broad frequency range to fit selected portions of seismograms. However, in practice, it requires substantial computational resources, often limiting the model domain size and resolution, dataset size, and frequency range. A more computationally efficient alternative is joint body and surface wave tomography, which simultaneously incorporates body-wave travel times and surface-wave dispersion measurements. This approach enables us to use more information in seismograms to constrain velocity structure. While simpler than the full-waveform inversion, it offers advantages such as complimentary sampling of Earth's interior by body and surface waves, enhanced spatial resolution with broader-frequency data, and improved computational efficiency.
Fang et al. (2015) developed a one-step inversion method for surface wave traveltimes at different frequencies without constructing the phase or group velocity maps, and Fang et al. (2016) conducted joint inversion of body and surface waves. In this study, to further reduce the gap between the joint inversion and full-waveform inversion, we incorporated finite-frequency sensitivity kernels for surface waves by Yoshizawa & Kennett (2004) into the formulation of Fang et al. (2015). For body wave data, we followed Obayashi et al. (2013) who employed a hybrid approach combining finite-frequency (Dahlen et al., 2000) and ray-theoretical kernels (Inoue et al., 1990).
We applied this method to construct the 3-D shear wave velocity structure beneath the Ontong Java Plateau (OJP) region. The broadband ocean-bottom seismic investigation, “OJP Array”, was conducted across OJP and adjacent areas from late 2014 to early 2017 (Suetsugu et al., 2018). The body-wave dataset was obtained from the OJP Array, including manually picked S-wave traveltimes and multiband differential traveltimes between two stations measured by cross-correlation. These were combined with S-wave traveltimes from the global catalog of ISC (International Seismology Centre, 2023). For surface wave data, we used Rayleigh and Love wave phase velocity dispersion measurements (Isse et al., 2021). We will present the resulting velocity structure in comparison with previous tomographic studies in the OJP region, both the multi-frequency tomography model using P-wave wave arrival times (Obayashi et al., 2021) and the radially anisotropic shear wave tomography model based on multi-mode surface wave dispersion measurements (Isse et al., 2021).
Fang et al. (2015) developed a one-step inversion method for surface wave traveltimes at different frequencies without constructing the phase or group velocity maps, and Fang et al. (2016) conducted joint inversion of body and surface waves. In this study, to further reduce the gap between the joint inversion and full-waveform inversion, we incorporated finite-frequency sensitivity kernels for surface waves by Yoshizawa & Kennett (2004) into the formulation of Fang et al. (2015). For body wave data, we followed Obayashi et al. (2013) who employed a hybrid approach combining finite-frequency (Dahlen et al., 2000) and ray-theoretical kernels (Inoue et al., 1990).
We applied this method to construct the 3-D shear wave velocity structure beneath the Ontong Java Plateau (OJP) region. The broadband ocean-bottom seismic investigation, “OJP Array”, was conducted across OJP and adjacent areas from late 2014 to early 2017 (Suetsugu et al., 2018). The body-wave dataset was obtained from the OJP Array, including manually picked S-wave traveltimes and multiband differential traveltimes between two stations measured by cross-correlation. These were combined with S-wave traveltimes from the global catalog of ISC (International Seismology Centre, 2023). For surface wave data, we used Rayleigh and Love wave phase velocity dispersion measurements (Isse et al., 2021). We will present the resulting velocity structure in comparison with previous tomographic studies in the OJP region, both the multi-frequency tomography model using P-wave wave arrival times (Obayashi et al., 2021) and the radially anisotropic shear wave tomography model based on multi-mode surface wave dispersion measurements (Isse et al., 2021).