11:00 AM - 1:00 PM
[SSS10-P08] An application of the seismic interferometry for observed records acquired by broadband seismometers installed around Tokyo Bay
Keywords:Seismic interferometry, Autocorrelation function, Seismic basement
- Outline -
In recent years, the seismic interferometry is applied in various regions and many studies have been already reported. Nakahara (2006) reported its theoretical background for one-dimensional layered media and it is derived that the Green's function corresponding to reflective waveform having surface source can be retrieved by the autocorrelation function of the ground surface observation record of earthquakes. For Kanto area, Yoshimoto et al. (2008) applied this method for dense seismic network and they presented excellent results which shows clear reflective phases from seismic basement.
Although their respectable results exists already, we also attempted to apply their methods to broadband seismic records installed in the thermal power plants by TEPCO. It is expected to complement observation areas where stations are relatively sparse in the previous studies.
- Analysis -
We used the waveforms observed by broadband velocity type seismometers (Tokyo sokushin: VSE355-G3) installed at thermal power plants by TEPCO mainly around Tokyo Bay area (Uetake, 2012). The relatively deep earthquakes which occurred in the northwestern Chiba prefecture and in the southern Ibaraki prefecture were selected to the analysis, because those incident angles is thought to be largely vertical. Twenty earthquakes with magnitude of 4.1 to 5.5 were used at the time of this abstract. The observed velocity waveforms were band-pass-filtered, and then were integrated to calculate displacement waveforms. After the calculation of the autocorrelation function of each transverse component waveform with a length of 15 s from the S-wave onset, these autocorrelation functions were stacked at each station. These stacked autocorrelation functions were assumed pseudo-reflectivity waveforms.
- Results -
At the observation stations located on the east side of Tokyo Bay, a relatively clear phase is observed at a position of about 7 seconds. These are almost the same as the time recognized as the reflected S wave phase from the seismic basement by Yoshimoto (2010). In addition, the time at each stations is also consistent with the two-way time for reflected S-wave, which is about 6 to 7 seconds, calculated using the subsurface structure model referring to JIVSM (Koketsu et al., 2012). On the other hand, even at the observation stations on the west side of Tokyo Bay, a phase with a negative amplitude can be seen at each stations around 6 to 8 seconds, which is interpreted as the reflected wave from the subsurface layer boundary. However, those are not always clear. It may can be understood that these differences reflect the variations between the eastern and western sides of Tokyo Bay in the subsurface structure, however, more detailed examination is required.
In addition, it is expected that a phases appeared about 0 to 4 seconds in two-way time reflect the information about relatively shallow layers, and those characteristics in pseudo-reflectivity waveforms at each observation station were different. Therefore, it may be possible to investigate the difference of shallow structures among the observation stations in Tokyo Bay area from obtained records.
Finally, it was difficult to distinguish reflective phases at some stations only from single pseudo-reflectivity waveforms. In contrast, the continuity of records around the objective station may assist to recognize the reflected phases easily and may enhance the reliability of picked phases. We would like to try such analyses.
In recent years, the seismic interferometry is applied in various regions and many studies have been already reported. Nakahara (2006) reported its theoretical background for one-dimensional layered media and it is derived that the Green's function corresponding to reflective waveform having surface source can be retrieved by the autocorrelation function of the ground surface observation record of earthquakes. For Kanto area, Yoshimoto et al. (2008) applied this method for dense seismic network and they presented excellent results which shows clear reflective phases from seismic basement.
Although their respectable results exists already, we also attempted to apply their methods to broadband seismic records installed in the thermal power plants by TEPCO. It is expected to complement observation areas where stations are relatively sparse in the previous studies.
- Analysis -
We used the waveforms observed by broadband velocity type seismometers (Tokyo sokushin: VSE355-G3) installed at thermal power plants by TEPCO mainly around Tokyo Bay area (Uetake, 2012). The relatively deep earthquakes which occurred in the northwestern Chiba prefecture and in the southern Ibaraki prefecture were selected to the analysis, because those incident angles is thought to be largely vertical. Twenty earthquakes with magnitude of 4.1 to 5.5 were used at the time of this abstract. The observed velocity waveforms were band-pass-filtered, and then were integrated to calculate displacement waveforms. After the calculation of the autocorrelation function of each transverse component waveform with a length of 15 s from the S-wave onset, these autocorrelation functions were stacked at each station. These stacked autocorrelation functions were assumed pseudo-reflectivity waveforms.
- Results -
At the observation stations located on the east side of Tokyo Bay, a relatively clear phase is observed at a position of about 7 seconds. These are almost the same as the time recognized as the reflected S wave phase from the seismic basement by Yoshimoto (2010). In addition, the time at each stations is also consistent with the two-way time for reflected S-wave, which is about 6 to 7 seconds, calculated using the subsurface structure model referring to JIVSM (Koketsu et al., 2012). On the other hand, even at the observation stations on the west side of Tokyo Bay, a phase with a negative amplitude can be seen at each stations around 6 to 8 seconds, which is interpreted as the reflected wave from the subsurface layer boundary. However, those are not always clear. It may can be understood that these differences reflect the variations between the eastern and western sides of Tokyo Bay in the subsurface structure, however, more detailed examination is required.
In addition, it is expected that a phases appeared about 0 to 4 seconds in two-way time reflect the information about relatively shallow layers, and those characteristics in pseudo-reflectivity waveforms at each observation station were different. Therefore, it may be possible to investigate the difference of shallow structures among the observation stations in Tokyo Bay area from obtained records.
Finally, it was difficult to distinguish reflective phases at some stations only from single pseudo-reflectivity waveforms. In contrast, the continuity of records around the objective station may assist to recognize the reflected phases easily and may enhance the reliability of picked phases. We would like to try such analyses.