Reference soil condition for intensity prediction equations derived from seismological and geophysical data at seismic stations

In 2011, an amplification map achieved by macroseismic information was developed for Switzerland using the collection of macroseismic intensity observations of past earthquakes. For each village, a ΔIm was first derived, which reflects the difference between observed and expected macroseismic intensities from a region-specific intensity prediction equation. The ΔIm values are then grouped into geological/tectonic classes, which are then presented in the macroseismic amplification map. Both, the intensity prediction equation and the macroseismic amplification map are referenced to the same reference soil condition which so far was only roughly estimated. This reference soil condition is assessed in this contribution using geophysical and seismological data collected by the Swiss Seismological Service. Geophysical data consist of shear-wave velocity profiles measured at the seismic stations and earthquake recordings, used to retrieve empirical amplification functions at the sensor locations. Amplification functions are referenced to a generic rock profile (Swiss reference rock condition) that is well defined, and it is used for the national seismic hazard maps. Macroseismic amplification factors Af, derived from empirical amplification functions, are assigned to each seismic station using ground motion to intensity conversions. We then assess the factors dΔf defined as the difference between Af and ΔIm. The factor dΔf accounts for the difference between the reference soil condition for the intensity prediction equation and the Swiss reference rock. We finally analysed relationships between Af and proxies for shear-wave velocity profiles in terms of average shear-wave velocity over defined depth ranges, such as VS,30, providing an estimate of the reference shear velocity for the intensity prediction equation and macroseismic amplification map. This study allows linking macroseismic intensity observations with experimental geophysical data, highlighting a good correspondence within the uncertainty range of macroseismic observations. However, statistical significance tests point out that the seismic stations are not evenly distributed among the various geological–tectonic classes of the macroseismic amplification map and its revision could be planned merging classes with similar behaviour or by defining a new classification scheme.