Revisiting Vs30 as a proxy parameter for site effects: a case study using KIKNET data

International audienceThe local geology can strongly affect the seismic ground motion at the surface. One standard index used to classify local site effects is the average shear wave velocity in the first 30 meters, namely Vs30. The higher the Vs30 the stiffer the material is and therefore less amplification is expected. However, this index does not map the complexity of the velocity at depth. Indeed, some sites may have close Vs30 values, but quite different velocity distribution at depth. In this study, we propose the ratio between the average shear wave velocity up to depths having at least 800 m/s and Vs30. This ratio gives information about the velocity distribution at depth. Lower ratios mean stronger similarity among velocity profiles. The proposed parameter is tested on the well-characterized Kik-net boreholes in Japan. These boreholes have P- and S-wave velocity profiles up to depths between 100 m and 200 m. Furthermore, we propose to study the empirical variability of site response using the recorded events at some sites. The combined use of Vs30 and the contrast ratio index allows a robust comparison of sites in order to evaluate the pertinence of Vs30 for site classification purposes. In order to avoid possible noonlinear site effects, only weak-motion data are used to compute linear site response (PGA \textless 20 cm/s/s)