Applicability of the HVSR method for fundamental frequency estimation of offshore sites: A theoretical investigation

Accurate characterization of offshore site conditions is critical for seismic hazard assessment, yet traditional methods based on shear-wave velocity ( V _S30) are often impractical and costly in marine environments. The Horizontal-to-Vertical Spectral Ratio (HVSR) method is a cost-effective technique widely used for estimating the fundamental frequency of onshore sites, but its effectiveness in offshore settings has not been validated. This study investigates the applicability and limitations of the HVSR method for offshore site characterization. We develop a one-dimensional (1D) offshore site model that incorporates the coupled dynamic effects of the overlying seawater layer and fluid-saturated soil layers (based on Biot's theory). Through a systematic parametric analysis, we compare HVSR curves with theoretical site transfer functions (as a benchmark) to evaluate the influence of seawater depth and seawater compressional wave (P-wave) velocity ( V _p). Our results reveal that the method's effectiveness is strongly governed by seawater depth. In deep-water environments (e.g., depths greater than 1000 m), the fundamental frequency identified from the HVSR peak shows excellent agreement with the benchmark. In contrast, under shallow-water conditions, the dynamic response of the seawater column introduces distinct spectral notches in the vertical spectrum that interfere with the HVSR curve, potentially leading to a significant misinterpretation of the true site frequency. This limitation under shallow-water conditions is confirmed by observed offshore data from S-net stations in Japan. These findings validate the HVSR method as a reliable and practical tool for site characterization in deep-water offshore settings, supporting its use in developing site-specific ground motion models. However, its application in shallow water requires extreme caution and methods to deconvolve the effects of seawater dynamics.