ABSTRACT
Based on the diffuse field concept for a horizontal‐to‐vertical spectral ratio of earthquakes (eHVSR), the effectiveness of eHVSRs to invert P‐ and S‐wave velocity structures down to the seismological bedrock (with the S‐wave velocity of or higher) has been shown in several published works. An empirical method to correct the difference between eHVSR and a horizontal‐to‐vertical ratio of microtremors (mHVSR), which is called earthquake‐to‐microtremor ratio (EMR), has also been proposed for strong‐motion sites in Japan. However, the applicability of EMR outside of Japan may not be warranted. We test EMR applicability for the Grenoble basin in France with plentiful microtremor data together with observed weak‐motion recordings at five sites. We thereby establish a systematic procedure to estimate the velocity structure from microtremors and delineate the fundamental characteristics of the velocity structures. We first calculate the EMR specific for the Grenoble basin () and calculate pseudo eHVSR (pHVSR) from and mHVSR. We compare the pHVSRs with the eHVSRs at five sites and find sufficient similarity to each other. Then, we invert velocity structures from eHVSRs, pHVSRs, and mHVSRs. The velocity structures from eHVSRs are much closer to those from pHVSRs than those from mHVSRs. We need to introduce a number of layers with gradually increasing S‐wave velocities below the geological basin boundary from a previous gravity study because the theoretical eHVSR of the model with a large velocity contrast has larger peak amplitudes than the observed. The depth of the S‐wave velocity of () shows a strong, linear correlation with the geological boundary depth. Finally, we apply our validated methodology and invert velocity structures using pHVSRs at 14 sites where there are no observed earthquakes. The overall picture of at a cross section in the northeastern part of the basin corresponds to the geological boundary.
