[HDS12-P07] The characteristic of ground motion amplification based on the Fourier spectrum ratio between surface and underground of the KiK-net seismic observation points - Basic study of landslides caused by earthquakes -
Keywords:Ground motion, KiK-net, Amplification property, Fourier spectrum, Transfer function
To evaluate ground motion amplification properties of earthquakes in the mountainous area is essential for a mechanism elucidation of landslides caused by earthquakes. Surface and underground seismometry records have been accumulated for more than 20 years after the 1995 Hyogo-ken-nambu Earthquake at the KiK-net observation points of the nationwide seismometry network. In the mountainous area, it is expected that the ground motion will be amplified by both the topography effect and by the ground effect. Nevertheless, the seismometry record of earthquakes in the mountainous area is hardly provided until now because there are very few observation points in the mountainous area.
A large number of small- and medium-scale earthquake data are provided by long-term observation of 14 KiK-net observation points located in mountainous areas. In addition, large earthquakes that triggered a large number of landslides can be successfully observed at some observation points in non-mountainous areas. The useful information about ground motion amplification properties of earthquakes in the mountainous area may be provided by combining these KiK-net observation data.
We collected acceleration records with PGA> 20 gal and analyzed the Fourier spectrum of ground motions (acceleration) at each KiK-net observation point by the following procedure.
1) For each earthquake, we calculated the Fourier spectrum of NS, EW, and UD components at the surface and underground.
2) For the surface and underground, we calculated the composition Fourier spectrum (the square root of the sum of squares of NS, EW, and UD components).
3) We calculated the ratio of the composition Fourier spectrum between the surface and underground for each frequency, which we call the amplification ratio spectrum. We then calculated the arithmetic mean of the amplification ratio spectrum of all earthquakes for each observation point, and call it the mean amplification ratio spectrum. We can consider it to be a kind of transfer function.
The following results were obtained.
1) The composition Fourier spectrum of the surface and underground changes according to the scale of earthquakes, but the amplification ratio spectrum does not change. The mean amplification ratio spectrum is considered to be an index in relation to the dominant frequency and amplification properties of each observation point.
2) The mean amplification ratio at frequency around 1 Hz is closed to 1, which means that there is no amplification.
3) The 14 KiK-net observation points at mountainous areas can be separated into three groups A, B, and C. In the case of group A, the dominant frequency (5-10 Hz) is relatively clear, and the maximum mean amplification ratio is more than 10, which means that the ground is easy to shake. In the case of group B, the dominant frequency is not clear, and the maximum mean amplification ratio is less than 10, which means that the ground is relatively difficult to shake.
These findings may provide important information on landslides caused by earthquakes.
A large number of small- and medium-scale earthquake data are provided by long-term observation of 14 KiK-net observation points located in mountainous areas. In addition, large earthquakes that triggered a large number of landslides can be successfully observed at some observation points in non-mountainous areas. The useful information about ground motion amplification properties of earthquakes in the mountainous area may be provided by combining these KiK-net observation data.
We collected acceleration records with PGA> 20 gal and analyzed the Fourier spectrum of ground motions (acceleration) at each KiK-net observation point by the following procedure.
1) For each earthquake, we calculated the Fourier spectrum of NS, EW, and UD components at the surface and underground.
2) For the surface and underground, we calculated the composition Fourier spectrum (the square root of the sum of squares of NS, EW, and UD components).
3) We calculated the ratio of the composition Fourier spectrum between the surface and underground for each frequency, which we call the amplification ratio spectrum. We then calculated the arithmetic mean of the amplification ratio spectrum of all earthquakes for each observation point, and call it the mean amplification ratio spectrum. We can consider it to be a kind of transfer function.
The following results were obtained.
1) The composition Fourier spectrum of the surface and underground changes according to the scale of earthquakes, but the amplification ratio spectrum does not change. The mean amplification ratio spectrum is considered to be an index in relation to the dominant frequency and amplification properties of each observation point.
2) The mean amplification ratio at frequency around 1 Hz is closed to 1, which means that there is no amplification.
3) The 14 KiK-net observation points at mountainous areas can be separated into three groups A, B, and C. In the case of group A, the dominant frequency (5-10 Hz) is relatively clear, and the maximum mean amplification ratio is more than 10, which means that the ground is easy to shake. In the case of group B, the dominant frequency is not clear, and the maximum mean amplification ratio is less than 10, which means that the ground is relatively difficult to shake.
These findings may provide important information on landslides caused by earthquakes.