2:30 PM - 2:45 PM
[SSS03-04] High-resolution CMT inversion for shallow offshore earthquake utilizing ocean-influenced Rayleigh wave thanks to thick seawater layer
Keywords:Centroid moment tensor, Offshore earthquake, Ocean-influenced Rayleigh wave
Determining high-resolution centroid moment tensor (CMT) solutions of offshore earthquakes is necessary for analyzing earthquake source mechanisms, conducting earthquake slip inversion, and predicting Tsunamis. However, for shallow earthquakes with focal depths less than 30 km, it is widely known that Mxz, Myz, and Mzz components are poorly constrained (Dziewonski et al., 1981; Kanamori & Given, 1981). These components are important for resolving dip angles, isotropic and CLVD components. Moreover, if Mxz, Myz, and Mzz components are predominate, focal depths cannot be determined with high-resolution.
We explored the potential of ocean-influenced Rayleigh waves (Noguchi et al., 2016), which are notably observed following the fundamental mode of continental Rayleigh waves for offshore earthquakes with thick (~4 km or more) seawater layers (e.g., Nakamura et al., 2012; Todoroki et al., 2017). They are not utilized in conventional CMT inversions based on 1-D seismic velocity structure model. Since these waves were remarkably observed for shallow offshore earthquakes, they have a potential to constrain CMT parameters of shallow offshore earthquakes. Still, the sensitivities of these waves to CMT parameters remain unclear.
In this study, we first investigated the characteristics of the ocean-influenced Rayleigh wave, focusing on their sensitivities for CMT parameters. We calculated the synthetic seismograms using 1-D seismic velocity structure models following Noguchi et al. (2016), with and without a 6 km seawater layer. We found that ocean-influenced Rayleigh waves have a larger amplitude in the vertical component compared to the radial component. The ocean-influenced Rayleigh wave exhibits higher sensitivity in Mxz, Myz, and Mzz components and focal depths mainly in the period range of 10–20 s. The amplitude and sensitivity decrease for longer period range, and finally disappear beyond 100 s.
Next, we conducted synthetic tests of CMT inversion using the vertical component, distributing seismic stations at 10 degrees azimuthal intervals at epicentral distances of 100, 200, and 300 km. We applied the band-pass filters in the period ranges of 10–20 s and 20–50 s. We also calculated Green’s functions with and without the seawater layer. By incorporating ocean-influenced Rayleigh wave, we obtained higher resolutions of focal depths compared to not utilizing the waves. Specifically using seismograms in the period range of 10–20 s, for earthquakes shallower than 20 km, the focal depths were constrained with ~10 km uncertainties, although those were with ~30 km without using the ocean-influenced Rayleigh wave. For earthquakes deeper than 20 km, CMT solutions with shallow depths can be rejected because the synthetic seismograms for shallower solutions cannot reproduce the target seismograms. The result suggests that Mxz, Myz, and Mzz components and focal depths are well constrained by using the ocean-influenced Rayleigh wave.
Finally, we applied this analysis to some outer-rise earthquakes in the Japan Trench. We calculated Green’s functions using a 3-D seismic velocity structure model by Koketsu et al. (2012) with topography and a seawater layer. We used the seismograms observed in the vertical components of the F-net stations. We affirm the high-resolution CMT solutions were constrained by utilizing the ocean-influenced Rayleigh wave. This approach has a significant advancement to constrain the CMT parameters.
We explored the potential of ocean-influenced Rayleigh waves (Noguchi et al., 2016), which are notably observed following the fundamental mode of continental Rayleigh waves for offshore earthquakes with thick (~4 km or more) seawater layers (e.g., Nakamura et al., 2012; Todoroki et al., 2017). They are not utilized in conventional CMT inversions based on 1-D seismic velocity structure model. Since these waves were remarkably observed for shallow offshore earthquakes, they have a potential to constrain CMT parameters of shallow offshore earthquakes. Still, the sensitivities of these waves to CMT parameters remain unclear.
In this study, we first investigated the characteristics of the ocean-influenced Rayleigh wave, focusing on their sensitivities for CMT parameters. We calculated the synthetic seismograms using 1-D seismic velocity structure models following Noguchi et al. (2016), with and without a 6 km seawater layer. We found that ocean-influenced Rayleigh waves have a larger amplitude in the vertical component compared to the radial component. The ocean-influenced Rayleigh wave exhibits higher sensitivity in Mxz, Myz, and Mzz components and focal depths mainly in the period range of 10–20 s. The amplitude and sensitivity decrease for longer period range, and finally disappear beyond 100 s.
Next, we conducted synthetic tests of CMT inversion using the vertical component, distributing seismic stations at 10 degrees azimuthal intervals at epicentral distances of 100, 200, and 300 km. We applied the band-pass filters in the period ranges of 10–20 s and 20–50 s. We also calculated Green’s functions with and without the seawater layer. By incorporating ocean-influenced Rayleigh wave, we obtained higher resolutions of focal depths compared to not utilizing the waves. Specifically using seismograms in the period range of 10–20 s, for earthquakes shallower than 20 km, the focal depths were constrained with ~10 km uncertainties, although those were with ~30 km without using the ocean-influenced Rayleigh wave. For earthquakes deeper than 20 km, CMT solutions with shallow depths can be rejected because the synthetic seismograms for shallower solutions cannot reproduce the target seismograms. The result suggests that Mxz, Myz, and Mzz components and focal depths are well constrained by using the ocean-influenced Rayleigh wave.
Finally, we applied this analysis to some outer-rise earthquakes in the Japan Trench. We calculated Green’s functions using a 3-D seismic velocity structure model by Koketsu et al. (2012) with topography and a seawater layer. We used the seismograms observed in the vertical components of the F-net stations. We affirm the high-resolution CMT solutions were constrained by utilizing the ocean-influenced Rayleigh wave. This approach has a significant advancement to constrain the CMT parameters.