5:15 PM - 7:15 PM
[STT42-P10] A comparison of acceleration and strain rate waveforms using wave propagation simulation
Keywords:distributed acoustic sensing, numerical simulation, wave propagation, strain rate, heterogeneous structure
DAS (Distributed Acoustic Sensing) is an ultra-dense seismic wave observation technology that uses fiber-optic cables as sensors. Strain along the fiber can be detected by transmitting an optical pulse from one end of the cable and measuring the phase change of its Rayleigh backscattered wave. DAS can measure strain or strain rate at intervals of tens of meters over tens of kilometers, enabling long and high linear density observation at low cost, which has been difficult with conventional seismometers. Another major advantage is that existing optical fiber communication infrastructure is available for DAS.
However, there are several difficulties in applying DAS to seismic observation. First, the strain observed by DAS corresponds to the spatial derivative of the displacement, velocity, and acceleration measured by conventional seismometers, making DAS sensitive to local changes in the topography and medium. Another difficulty is that DAS can measure only the axial strain of a fiber. This means that the sensitivity of the DAS varies considerably depending on the incident angle of the seismic waves.
Numerical simulations are useful for verifying how the abovementioned characteristics affect actual observational results. For example, Capdeville et al. (2024) numerically analyzed strain waveforms in regions where heterogeneities exist at scales much smaller than the minimum wavelength of the wavefield using the spectral element method. Their results suggest that DAS may exhibit high sensitivity to such small-scale heterogeneities.
In this study, we investigated the difference between the acceleration and strain rate waveforms of a plane P wave passing through a low-velocity domain compared to the surrounding area using numerical simulations.
We consider the two-dimensional elastic medium of size 20 × 40 km2 (VP = 4.4 km/s, VS = 2.7 km/s, ρ = 2700 kg/m3 ), in which 8 × 8 km2 square low-velocity domain (VP = 2.4 km/s, VS = 1.5 km/s, ρ = 1500 kg/m3 ) is embedded. The wave was excited as a Ricker wavelet with a center frequency of 1 Hz and was driven by solving the two-dimensional elastodynamic equations by the finite difference method (Virieux, 1984). Mimicking a DAS, acceleration and strain rate data were collected on a straight line traversing a low-velocity domain.
The results showed that in the low velocity region, the acceleration was significantly amplified away from the boundary, while the strain rate was significantly amplified on and near the boundary. This difference is caused by a polarity reversal of either the acceleration or the strain rate when the wave is reflected at the boundary.The results also show a discrepancy between the acceleration and distortion velocity waveforms due to diffracted waves. If the wavefield consists of a single plane wave, the acceleration and strain rate waveforms are simply proportional, but this condition is no longer valid due to the presence of a low-velocity domain.
However, there are several difficulties in applying DAS to seismic observation. First, the strain observed by DAS corresponds to the spatial derivative of the displacement, velocity, and acceleration measured by conventional seismometers, making DAS sensitive to local changes in the topography and medium. Another difficulty is that DAS can measure only the axial strain of a fiber. This means that the sensitivity of the DAS varies considerably depending on the incident angle of the seismic waves.
Numerical simulations are useful for verifying how the abovementioned characteristics affect actual observational results. For example, Capdeville et al. (2024) numerically analyzed strain waveforms in regions where heterogeneities exist at scales much smaller than the minimum wavelength of the wavefield using the spectral element method. Their results suggest that DAS may exhibit high sensitivity to such small-scale heterogeneities.
In this study, we investigated the difference between the acceleration and strain rate waveforms of a plane P wave passing through a low-velocity domain compared to the surrounding area using numerical simulations.
We consider the two-dimensional elastic medium of size 20 × 40 km2 (VP = 4.4 km/s, VS = 2.7 km/s, ρ = 2700 kg/m3 ), in which 8 × 8 km2 square low-velocity domain (VP = 2.4 km/s, VS = 1.5 km/s, ρ = 1500 kg/m3 ) is embedded. The wave was excited as a Ricker wavelet with a center frequency of 1 Hz and was driven by solving the two-dimensional elastodynamic equations by the finite difference method (Virieux, 1984). Mimicking a DAS, acceleration and strain rate data were collected on a straight line traversing a low-velocity domain.
The results showed that in the low velocity region, the acceleration was significantly amplified away from the boundary, while the strain rate was significantly amplified on and near the boundary. This difference is caused by a polarity reversal of either the acceleration or the strain rate when the wave is reflected at the boundary.The results also show a discrepancy between the acceleration and distortion velocity waveforms due to diffracted waves. If the wavefield consists of a single plane wave, the acceleration and strain rate waveforms are simply proportional, but this condition is no longer valid due to the presence of a low-velocity domain.