scholarly journals Temporal change in seismic wave attenuation using highly stable vibration sources

2022 ◽  
Vol 74 (1) ◽  
Author(s):  
Shuhei Tsuji ◽  
Koshun Yamaoka ◽  
Ryoya Ikuta

AbstractWe developed a method to detect attenuation changes during seismic wave propagation excited by precisely controlled artificial seismic sources, namely Accurately Controlled Routinely Operated Signal System (ACROSS), and applied it to monitor the temporal changes for in situ data collected by previous studies. Our method, together with the use of the ACROSS sources, is less susceptible to noise level changes, from which conventional methods such as envelope calculation suffer. The method utilizes the noise level that is independently estimated in the frequency domain and eliminates the influence of the noise from the observed signal. For performance testing, we applied this method to a dataset that was obtained in an experiment at Awaji Island, Japan, from 2000 to 2001. We detected a change in amplitude caused by rainfall, variation in atmospheric temperature, and coseismic ground motions. Among them, coseismic changes are of particular interest because there are limited studies on coseismic attenuation change, in contrast to many studies on coseismic velocity decrease. At the 2000 Western Tottori earthquake (MW = 6.6, epicenter distance of 165 km), a sudden decrease in amplitude of up to 5% was observed. The coseismic amplitude reduction and its anisotropic characteristics, which showed a larger reduction in the direction of the major axis of velocity decrease, were consistent with the opening of fluid-filled cracks, as proposed by previous studies. The $$\Delta {Q}^{-1}$$ Δ Q - 1 corresponding to the amplitude change gives similar values to those reported in previous studies using natural earthquakes. Graphical Abstract

2021 ◽  
Author(s):  
Shuhei TSUJI ◽  
Koshun Yamaoka ◽  
Ryoya Ikuta

Abstract We developed a method to detect attenuation changes during seismic wave propagation excited by precisely controlled artificial seismic sources, namely Accurately Controlled Routinely Operated Signal System (ACROSS), and applied it to monitor the temporal changes for in situ data collected by previous studies. Our method, together with the use of the ACROSS sources, is less susceptible to noise level changes, from which conventional methods such as envelope calculation suffer. The method utilizes the noise level that is independently estimated in the frequency domain and eliminates the influence of the noise from the observed signal. For performance testing, we applied this method to a dataset that was obtained in an experiment at Awaji Island, Japan, from 2000 to 2001. We detected a change in amplitude caused by rainfall, variation in atmospheric temperature, and coseismic ground motions. Among them, coseismic changes are of particular interest because there are limited studies on coseismic attenuation change, in contrast to many studies on coseismic velocity decrease. At the 2000 Western Tottori earthquake (MW = 6.6, epicenter distance of 165 km), a sudden decrease in amplitude of up to 5% was observed. The coseismic amplitude reduction and its anisotropic characteristics, which showed a larger reduction in the direction of the major axis of velocity decrease, were consistent with the opening of fluid-filled cracks, as proposed by previous studies. The W corresponding to the amplitude change gives similar values to those reported in previous studies using natural earthquakes.


2020 ◽  
Author(s):  
Shuhei Tsuji ◽  
Koshun Yamaoka ◽  
Ryoya Ikuta

Abstract We developed a method to detect changes in attenuation in transfer functions obtained by precisely controlled artificial seismic sources, namely Accurately Controlled Routinely Operated Signal System (ACROSS), and applied it to monitor the temporal changes for in-situ data collected by previous studies. Our method, together with the usage of the ACROSS sources, is less susceptible to change in noise level to which conventional methods such as envelope calculation suffer. The method utilizes the noise level that is independently estimated in the frequency domain. Thus, we can eliminate the influence of the noise by subtracting it from the observed signal. To test the performance, we applied this method to a dataset obtained at Awaji Island, Japan from 2000 to 2001. We detected the changes in amplitude with several causes including ground motion in an earthquake and water injection experiment. At the 2000 Western Tottori earthquake (MW = 6.6, Epicenter distance of 165 km), a sudden decrease in amplitude up to 5% followed by gradual recovery are clearly observed. These coseismic changes in amplitude are consistent with the opening of fluid-filled cracks as proposed by previous studies. We convert the amplitude change into ΔQ-1, which gives similar values as reported by previous studies using natural earthquakes. Increases in amplitude up to 5% associated with water injection experiments are also observed. During these experiments, the amplitude increased several days after the beginning of the injection and recovered to the previous level. This may be the result of a stress increase caused by the injection followed by a saturation increase by water diffusion.


Sign in / Sign up

Export Citation Format

Share Document