5:15 PM - 6:45 PM
[STT36-P09] Relationship between strain detected by DAS observation at Sakurajima and surface geology
Keywords:DAS (Distributed Acoustic Sensing), Site amplification factor, Sakurajima, Surface geology
Introduction
The observed seismic wave amplitudes are affected by the site amplification just beneath the receiver and we need to correct this effect by using the coda normalization method, for example. The site amplification is also essential for the prediction of the strong ground motion; therefore, high-resolution mapping of the site amplification is required. In Japan, we have 250 m mesh ground information, but higher resolution is still required. For estimation the site amplification, it is necessary to separate source and path effect from seismic wave observed by many observation points, but it is tough about the cost. The massive cost of the observation is necessary in the conventional method using many seismometers and observation points to obtain high-resolution data. In this study, we estimate site amplification characteristics using spatially dense seismic data obtained along the cables by DAS (Distributed Acoustic Sensing) observations using optical fiber cables, recently introduced into geophysics. In this presentation, we will discuss the relationship between high-resolution site amplification characteristics (spatial variation of amplitude) obtained from DAS observation data using relatively simple methods and the nature of the wave field, topography and geology, and verify the methods.
Data and method
From the DAS observation in 2022 using cables along the road encircling Sakurajima, 0-3000 channels (mainly in the south) with good signal-to-noise ratios were used. The measurements are strain time series with 4 m interval and 200 Hz time sampling. After excluding sections with local coupling problems and traffic noise, and after applying a bandpass filter of 1~2 Hz to 5 seismic events, the average RMS amplitude for the time window of 30 s including S-wave components for each channel. We also calculate the background noise level during the quiet period from 2 to 3 am averaged over 18 days. The averaged amplitudes were plotted on geologic and topographic maps.
Result
Spatial difference of amplitude has 100x range that is approximately high around 0~1500ch and 2100~800ch, and low around 1500~2100ch and 2800~3000ch, 5 seismic events and background noise level have similar trend. We use the average of each 20ch for mapping, so we recognize horizontal distribution of signal amplification approximately each 100m along the cable. The geological features are characterized broadly two sections that first is old and second is new lava flow. First section includes alluvial fan around 500ch, we recognize local high amplification in boundary between this small section and old lava flow. Around 2800~3000ch, which have boundary of Taisho-lava flow and other old lava flow, are also estimated higher amplification than before 2800ch.
Discussion
In the alluvial fan near 500ch, the difference in seismic motions given by lava and other geological features suggests that the geological boundary affects the amplification of the DAS strain record. The amplification rate on the lava flow away from the geologic boundary is low, and it can be evaluated that the ground is strong compared to the sediments. However, the Taisho lava flow 2200~2800ch has a low amplification rate, while 1500~2200ch, same geological features, has a high amplification rate, which contradicts the inference that lava is more solid and less prone to shaking. Further analyses comparing with seismometers are necessary to separate the effect of the cable coupling and the shallow structure. It is interesting to note that although there is a significant age gap of more than 300 years, geologically the lava flows are known to be andesitic and have large signal amplification compared to other lava flows such as the first half of the Taisho lava flow described above. Since the age of the lava flow is new, some parts of the lava flow are relatively collapsed and stable, while others are not. The wrinkle pattern of the lava flow differs between the first and second half of the Taisho lava, which may reflect the topographical difference indicating differences in solidification.
The observed seismic wave amplitudes are affected by the site amplification just beneath the receiver and we need to correct this effect by using the coda normalization method, for example. The site amplification is also essential for the prediction of the strong ground motion; therefore, high-resolution mapping of the site amplification is required. In Japan, we have 250 m mesh ground information, but higher resolution is still required. For estimation the site amplification, it is necessary to separate source and path effect from seismic wave observed by many observation points, but it is tough about the cost. The massive cost of the observation is necessary in the conventional method using many seismometers and observation points to obtain high-resolution data. In this study, we estimate site amplification characteristics using spatially dense seismic data obtained along the cables by DAS (Distributed Acoustic Sensing) observations using optical fiber cables, recently introduced into geophysics. In this presentation, we will discuss the relationship between high-resolution site amplification characteristics (spatial variation of amplitude) obtained from DAS observation data using relatively simple methods and the nature of the wave field, topography and geology, and verify the methods.
Data and method
From the DAS observation in 2022 using cables along the road encircling Sakurajima, 0-3000 channels (mainly in the south) with good signal-to-noise ratios were used. The measurements are strain time series with 4 m interval and 200 Hz time sampling. After excluding sections with local coupling problems and traffic noise, and after applying a bandpass filter of 1~2 Hz to 5 seismic events, the average RMS amplitude for the time window of 30 s including S-wave components for each channel. We also calculate the background noise level during the quiet period from 2 to 3 am averaged over 18 days. The averaged amplitudes were plotted on geologic and topographic maps.
Result
Spatial difference of amplitude has 100x range that is approximately high around 0~1500ch and 2100~800ch, and low around 1500~2100ch and 2800~3000ch, 5 seismic events and background noise level have similar trend. We use the average of each 20ch for mapping, so we recognize horizontal distribution of signal amplification approximately each 100m along the cable. The geological features are characterized broadly two sections that first is old and second is new lava flow. First section includes alluvial fan around 500ch, we recognize local high amplification in boundary between this small section and old lava flow. Around 2800~3000ch, which have boundary of Taisho-lava flow and other old lava flow, are also estimated higher amplification than before 2800ch.
Discussion
In the alluvial fan near 500ch, the difference in seismic motions given by lava and other geological features suggests that the geological boundary affects the amplification of the DAS strain record. The amplification rate on the lava flow away from the geologic boundary is low, and it can be evaluated that the ground is strong compared to the sediments. However, the Taisho lava flow 2200~2800ch has a low amplification rate, while 1500~2200ch, same geological features, has a high amplification rate, which contradicts the inference that lava is more solid and less prone to shaking. Further analyses comparing with seismometers are necessary to separate the effect of the cable coupling and the shallow structure. It is interesting to note that although there is a significant age gap of more than 300 years, geologically the lava flows are known to be andesitic and have large signal amplification compared to other lava flows such as the first half of the Taisho lava flow described above. Since the age of the lava flow is new, some parts of the lava flow are relatively collapsed and stable, while others are not. The wrinkle pattern of the lava flow differs between the first and second half of the Taisho lava, which may reflect the topographical difference indicating differences in solidification.