A COMPARISON OF SITE SHEAR WAVE VELOCITY PROFILE USING TWO METHODS OF INVERSION OF MAXIMUM ELLIPTICITY AND SPECTRAL ANALYSIS OF SURFACE WAVES, CASE

Shear wave velocity profile of construction site is one of the most important parameters that is requested for evaluation of seismic strength of sub-surface layers and their effects on the seismic design of surface structures known as site effects. There are several known methods to extract the shear wave velocity profile of ground. Amongst, some methods were developed based on analysis of surface waves from active or passive sources. This study focuses on the comparison of two of these methods; the inversion of maximum ellipticity of Rayleigh wave and the spectral analysis of surface wave (SASW). The former is based on the microtremor measurements of ambient vibrations in three components and construction of the spectral ratio of horizontal to vertical components, and the latter, is based on the recording of phase delay between vertical vibrations of ground level due to an active source like hammer impact. In this study, microtremor measurements at 21 stations of Kermanshah city, which has
been done in the project of world bank 4697-IRN, are used to determine the shear wave velocity profile using inversion of their ellipticity peak. Accordingly, shear wave velocity profile, are determined by using four models of 3, 5, 7 and 10 layers and finally the best shear wave velocity profiles resulted from inversion of microtremor ellipticity peak are compared with the results of spectral analysis of surface waves which were obtained at the same stations. The importance of current study is that the initial shear wave velocity profiles are suggested without considering the results of SASW, for considering the thickness and velocity limit of each layer. Comparisons reveal that good correlations are obtained between shear wave profiles of current study and results of SASW. Furthermore, the similarities of shear wave profiles obtained by two methods improve with increasing the number of considered layers of initial models from 3 to 10.