A STUDY ON MODELING OF DEEP SUBSURFACE STRUCTURE CONSIDERING RANDOM VELOCITY FLUCTUATION FOR ESTIMATION OF BROADBAND SITE AMPLIFICATION FACTOR

Abstract

The site amplification factors in Niigata plain, Japan, are empirically estimated by the spectral inversion analysis from S-wave parts of strong motion data. The resultant characteristics of the site amplification are relative large in the frequency range of about 1.5-5 Hz, and decay significantly with the frequency increasing over about 5 Hz. In order to explain these features, we propose to consider random fluctuation in the modeling of deep S-wave velocity structure for site factor estimation. As a result, the site amplification in the frequency range higher than about 5Hz are filtered out by the effects of the velocity fluctuation similar to those observed in Niigata plain. We further demonstrate that the addition of the damping factor in subsurface modeling also has the same high-cut filtering effect on the site amplification factor estimation. We also discuss the relationship between the additional damping factor and the strength of random fluctuation.