Seismic hazard studies provide indicators of seismic motion that are expressed for “free‐field,” that is, representative of the ground motion exactly at the free surface, without disturbances due to interactions between soil and buildings or other structures. Most of these studies are based on ground‐motion prediction equations, which are, themselves, formulated to predict free‐field motion, as they are derived from similarly free data. However, is this really the case?
In this study, we use several examples to illustrate how small structures hosting permanent strong‐motion stations (often anchored on small concrete slabs) generate soil–structure interaction effects that can amplify the high‐frequency part of the earthquake signal () by up to a factor of 2–3 for stations on soils. We also show that the installation depth of a station, even if very shallow (i.e., a few meters), can change the recorded response, mainly by deamplifying the signal in high frequencies () by a factor up to 0.3. Such effects imply that there are actual differences between recorded and true free‐field signals. Depending on the housing conditions, these effects can have significant impact on response spectra at high frequencies, and on measurements of the parameter.
It is, thus, becoming clear that such effects should be taken into account in studies involving high‐frequency seismic motion. To do so, scientists need a detailed description of the conditions of installation and housing of seismological and accelerometric stations, which often lacks from the metadata distributed through the various, commonly used web services. Increasing such information and facilitating the access to it would allow the identification of stations that are problematic and of those that are truly close to free‐field recording conditions. In a subsequent step, it would be important to quantify the modification curve of the response of stations that experience such effects.