An algorithm for identifying the surface waves, Love- and Rayleigh-waves,
in microtremors has already been developed as a new exploration
technique which can determine S-wave velocity structure. A practical
use of the technique has been designed and also examined at two
experimental sites in an urban area, Morioka city in northern
Honshu, Japan. The technique involves the measurement of the dispersion
of both surface waves in microtremors by using a specially-designed
circular seismic array consisting of more than three tri-axial
seismometers that are equally spaced on a circle and one at the
circular center. In determination of phase velocities of Love-
and Rayleigh- waves, the newly-modified spatial auto-correlation
method has been used. The method can provide not only the dispersion
of Rayleigh waves, but also that of Love waves. The dispersion
of Love waves is directly convertible to the S-wave velocity structure.
Estimated phase velocities of Love wave using microtremors at
the experiment site of Iwate University were in the range of 750
to 200 m/s corresponding to the frequency range from 2.8 to 10
Hz. The phase velocities determined from horizontal microtremors
were in good agreement with those calculated by the structure
model which were derived from the phase velocities of Rayleigh
waves. Estimated phase velocities of Love wave in microtremors
were in the range of 450 to 200 m/s corresponding to the frequency
range from 1 to 2.9 Hz at the experiment site of Morioka technical
high-school. These results of two experimental sites show that
the newly-modified spatial auto-correlation method is applicable
to estimation of Love wave phase velocities using microtremors.
Key Words: microtremor array observation, dispersion curve of surface waves, Love wave, Rayleigh wave, spatial auto-correlation method, S-wave velocity structure.
Spatial autocorrelation method(SAC) is an effective analysis for
estimating underground S-wave velocity structure from microtremor
phase velocity dispersion relation because it has larger detectable
range of microtremor wavelength than frequency - wavenumber analysis.
However, phase velocities estimated by conventional SAC methods
such as band-pass filtered method or Fast Fourier Transform method
were not precise if suitable band width was not selected for analysis.
We proposed a new technique for SAC using autoregressive model
which estimated spectra with high resolution because the best
fitting model can be selected using AIC. We apply the new method
to calculate phase velocities of microtremors which were observed
at a ground of Morioka Technical High School with arrays. As a
result, phase velocities calculated by the new method were continous
with frequency although those calculated by the conventional methods
were scattered. This indicates that SAC functions calculated by
the new method are estimated better than those by conventional
SAC methods .
Key words: spatial autocorrelation method, microtremor
array observation, autoregressive model, S-wave velocity structure
estimation
It is important to know underground velocity structure, especially
shallow S-wave velocity structure, in urban areas to study on
seismic microzoning or to predict strong motion in the high frequency
range during a large earthquake. However, it is difficult to carry
out a survey such as refraction or reflection because there is
no space for a survey. Microtremors exist anytime or anywhere
in urban areas. They are thought to construct an ensemble with
body and surface waves. If we can detect phase velocities of surface
wave from microtremors, we can estimate underground velocity structure
from the dispersion relation using an inversion technique. We
carried out array observations of short-period microtremors at
eleven sites in Morioka area to detect phase velocities as a function
of frequency using frequency-wavenumber analysis. Then, we estimated
underground velocity structure which consists of some layers on
half-space from the dispersion relation of Raylaigh wave, using
a linearized inversion technique. The shallow underground velocity
structures were estimated from Rayleigh wave dispersion relations
at nine sites. They were consistent with S-wave velocity structures
obtained from bore-hole data. It was shown that short-period microtremor
array observation was useful for prospecting shallow structure
beneath urban areas where there is no space for other geophysical
prospecting technique.
Keywords: short-period microtremor array observation, Rayleigh
wave dispersion relation, frequency-wavenumber analysis, inversion
of S-wave velocity structure
S-wave velocity information plays a major role in the study
of seismic micro-zoning. An empirical equation for estimating
S-wave velocity based on ground information obtained from borehole
data in Morioka city area in Tohoku district was constructed.
Refraction survey using the plank-hammering technique was carried
out at 76 points near borehole sites in order to obtain S-wave
velocity data. The total number of data used for multiple regression
analysis was 460. Seven empirical equations were estimated using
N-value, depth and facies as predictor variables. The best equation
obtained by AIC test is as follows:
Vs = 92.90 H**0.117 N**0.251 (1.000, 1.101, 1.153, 1.444)F
Vs: S-wave velocity(m/s)
H: depth(m)
N: N-value
F: Facies (clay, silt, sand, sand&gravel)
The correlation coefficient was 0.899. Although the calculated
velocities using the equations from the previous studies were
lower than the observed values in Morioka area, the calculated
velocities using the equation in the present study were nearly
equal to the observed values. This suggests that the above equation
reflects geological features in Morioka area.
Key words: empirical equation, S-wave velocity, borehole
data, plank-hammering technique.