11:30 AM - 11:45 AM
[STT42-08] Seismic observation with DAS using optical fiber cable along a Kyushu Shinkansen railway
Keywords:DAS, earthquake, railway
In recent years, the implementation of distributed acoustic sensing (DAS) for seismological purposes has rapidly expanded (e.g., Miah and Potter 2017, Ajo-Franklin et al., 2019, Sladen et al., 2019, Lindsey et al. 2020, Walter et al., 2020). DAS measures changes in strain along an optical fiber cable by using variations of the phase of backscattered laser pulses traveling in the optical fiber cable. The spatial resolution is 0.2-20 m; therefore, a few tens of meters of seismometers can be used for linear array observations corresponding to the length of the fiber cable. However, the obtained data, especially the amplitude information, is different from that obtained by conventional seismometers, and differs considerably depending on the installation environment of the optical fiber cables. In this study, we proposed conducting DAS seismic observations using fiber-optic cables along the railway. We applied DAS to the optical fiber cable along the Kyushu Shinkansen line operated by the Kyushu Railway Company in Kumamoto prefecture, where aftershocks from the 2016 Mw7.3 Kumamoto earthquake are still highly active.
We successfully observed strong motions of the Mj6.6 earthquake on January 22, 2022 in Hyuga-nada and several small local earthquakes for distances over 75 km. Both large P and S waves are visible in most of the channels, except for uncoupled channels with small signal levels (tunnel). A accelerometer that simultaneously observed the event recorded a maximum of approximately 80 gal.
The shaking map (maximum strain distributions) for Mj6.6 was estimated by correcting the cycle skipping due to the dynamic range. The differential phase data indicated cycle skipping at various channels, which saturated the data because waveform amplitude reached near the maximum of the dynamic range (−π to π in phase, which is linearly scaled between 32768 and -32768 (±215)) (Hartog, 2017; Ide et al., 2021). We estimated the data of the cycle-skipped channels using data from adjacent channels that were not cycle-skipped.
The attenuation property of local earthquakes was identified by correcting the site effect, coupling and amplification of the seismic wave by the railway structure. The peak strain value is estimated against the hypocentral distance to analyze the several local earthquakes in Kumamoto prefecture. In addition, the peak ground velocity (PGV) of NS component of each Hi-net station is estimated against the hypocentral distance. We observed that the peak strain values decreased with increasing hypocenter distance for each magnitude category as well as PGV. Furthermore, estimating the seismic attenuation using DAS is possible above the hypocenter (~10 km), wherein the number of data are inevitably reduced by using permanent seismometers. Our result indicates that the accurately corrected DAS amplitude value can empirically estimate the earthquake magnitude. Thus, the scaling relations obtained in this study have great potential for earthquake early warning.
We successfully observed strong motions of the Mj6.6 earthquake on January 22, 2022 in Hyuga-nada and several small local earthquakes for distances over 75 km. Both large P and S waves are visible in most of the channels, except for uncoupled channels with small signal levels (tunnel). A accelerometer that simultaneously observed the event recorded a maximum of approximately 80 gal.
The shaking map (maximum strain distributions) for Mj6.6 was estimated by correcting the cycle skipping due to the dynamic range. The differential phase data indicated cycle skipping at various channels, which saturated the data because waveform amplitude reached near the maximum of the dynamic range (−π to π in phase, which is linearly scaled between 32768 and -32768 (±215)) (Hartog, 2017; Ide et al., 2021). We estimated the data of the cycle-skipped channels using data from adjacent channels that were not cycle-skipped.
The attenuation property of local earthquakes was identified by correcting the site effect, coupling and amplification of the seismic wave by the railway structure. The peak strain value is estimated against the hypocentral distance to analyze the several local earthquakes in Kumamoto prefecture. In addition, the peak ground velocity (PGV) of NS component of each Hi-net station is estimated against the hypocentral distance. We observed that the peak strain values decreased with increasing hypocenter distance for each magnitude category as well as PGV. Furthermore, estimating the seismic attenuation using DAS is possible above the hypocenter (~10 km), wherein the number of data are inevitably reduced by using permanent seismometers. Our result indicates that the accurately corrected DAS amplitude value can empirically estimate the earthquake magnitude. Thus, the scaling relations obtained in this study have great potential for earthquake early warning.