HR: 1340h
AN: S33A-0299    [Abstracts]
TI: A Two-Radius Circular Array Method: Extracting Independent Information on Phase Velocities of Love Waves From Microtremor Records From a Simple Seismic Array
AU: * Tada, T
EM: kogutek@rs.kagu.tus.ac.jp
AF: Tokyo University of Science, 1-3 Kagurazaka, Shinjuku-ku, Tokyo, 162-8601 Japan
AU: Cho, I
EM: ikuo-chou@aist.go.jp
AF: Geological Survey of Japan, AIST, Central 7, 1-1-1 Higashi, Tsukuba, 305-8567 Japan
AU: Shinozaki, Y
EM: sinozaki@rs.kagu.tus.ac.jp
AF: Tokyo University of Science, 1-3 Kagurazaka, Shinjuku-ku, Tokyo, 162-8601 Japan
AB: We have invented a Two-Radius (TR) circular array method of microtremor exploration, an algorithm that enables to estimate phase velocities of Love waves by analyzing horizontal-component records of microtremors that are obtained with an array of seismic sensors placed around circumferences of two different radii. The data recording may be done either simultaneously around the two circles or in two separate sessions with sensors distributed around each circle. Both Rayleigh and Love waves are present in the horizontal components of microtremors, but in the data processing of our TR method, all information on the Rayleigh waves ends up cancelled out, and information on the Love waves alone are left to be analyzed. Also, unlike the popularly used frequency-wavenumber spectral (F-K) method, our TR method does not resolve individual plane-wave components arriving from different directions and analyze their "vector" phase velocities, but instead directly evaluates their "scalar" phase velocities --- phase velocities that contain no information on the arrival direction of waves --- through a mathematical procedure which involves azimuthal averaging. The latter feature leads us to expect that, with our TR method, it is possible to conduct phase velocity analysis with smaller numbers of sensors, with higher stability, and up to longer-wavelength ranges than with the F-K method. With a view to investigating the capabilities and limitations of our TR method in practical implementation to real data, we have deployed circular seismic arrays of different sizes at a test site in Japan where the underground structure is well documented through geophysical exploration. Ten seismic sensors were placed equidistantly around two circumferences, five around each circle, with varying combinations of radii ranging from several meters to several tens of meters, and simultaneous records of microtremors around circles of two different radii were analyzed with our TR method to produce estimates for the phase velocities of Love waves. The estimates were then checked against "model" phase velocities that are derived from theoretical calculations. We have also conducted a check of the estimated spectral ratios against the "model" spectral ratios, where we mean by "spectral ratio" an intermediary quantity that is calculated from observed records prior to the estimation of the phase velocity in the data analysis procedure of our TR method. In most cases, the estimated phase velocities coincided well with the model phase velocities within a wavelength range extending roughly from 3r to 6r (r: array radius). It was found out that, outside the upper and lower resolution limits of the TR method, the discrepancy between the estimated and model phase velocities, as well as the discrepancy between the estimated and model spectral ratios, were accounted for satisfactorily by theoretical consideration of three factors: the presence of higher surface-wave modes, directional aliasing effects related to the finite number of sensors in the seismic array, and the presence of incoherent noise.
DE: 0935 Seismic methods (3025, 7294)
DE: 7212 Earthquake ground motions and engineering seismology
DE: 7255 Surface waves and free oscillations
SC: Seismology [S]
MN: Fall Meeting 2005