Preliminary Analysis for Evaluation of Local Site Effects from Strong Motion Spectra by an Inversion Method

Abstract

Simultaneous separation of source, propagation path, and local site effects from observed strong motion records is carried out by an inversion method in the frequency range from 1 to 10 Hz for the purpose of empirical evaluation of the local site effects in different geological conditions. The analyzed data are S-wave portions of 167 accelerograms from 20 events observed at 7 stations along the Pacific coast of southern Tohoku and Kanto districts, Japan. These events are shallow earthquakes with magnitude from 4.0 to 6.7 and with hypocentral distance from 43 to 243 km. A linear inversion method is applied to the logarithm of the observed spectra, and solutions for source spectra, inelasticity factor of propagation path for S-waves (Q_s-value), and factor of site amplification at each station are obtained in least squares sense. In the inversion, the factor of site amplification at a reference station (KDG) on an outcrop of bedrock is constrained to 2. The factors of site amplification Gj(f) obtained in the present study at two borehole stations are verified by comparisons with the theoretical transfer functions from bedrock to observation points. Obtained Gj(f) values increase with decrease of shear-wave velocities of the ground layers at the observation points. Q_s-value obtained shows frequency dependency of the form Q_s(f)_??_60f_1.0, and amplitudes of displacement source spectra have ω-square decay in the frequency range higher than the corner frequency. The site effect at the reference station (KDG) is also discussed in detail to examine the validity of the constraint condition adopted in the present study. On the basis of the mean acceleration spectrum obtained from the observed records and the calculated topographical effects by the two-dimensional BEM model at KDG, it is concluded that the constraint condition is satisfactory in the frequency range lower than 6 Hz. In the range higher than 6 Hz, it is suggested that KDG is influenced by the topographical effects.