A REGRESSION MODEL FOR ESTIMATING POWER SPECTRAL DENSITY FUNCTION OF GROUND ACCELERATION

摘要

The purpose of this study is to develop a new regression model for power spectral density function (PSDF) of ground acceleration by using the strong motion data of the K-NET and the KiK-NET in Japan. The PSDF is simply expressed as the PA model, proposed by authors before, by using three parameters: the mean square value of the ground acceleration, the predominant frequency, and the shape factor. The regression analyses for the PA-model parameters are carried out based on the maximum likelihood method with respect to the explanatory variables: the distance from the fault to the site, the magnitude, the focal depth, and the averaged shear-wave velocity at the site. The obtained PSDF regression model can express well the averaged tendency, in quality and in quantity, of the observed PSDF. Furthermore, the efficiency and the applicability of the proposed model are demonstrated through the estimation of the earthquake input energy based on the random vibration theory.