Improved analysis of horizontal-to-vertical spectral ratio measurements for groundwater investigations

Abstract

Ambient seismic noise, and specifically the Horizontal-to-Vertical Spectral Ratio (HVSR), is routinely used for seismic microzonation, assessment of earthquake site characteristics and bedrock depth information for hydrogeological studies. These measurements not only provide peak frequency or period of a seismic resonator, but the shape of the spectral ratio can also give insight into the architecture of subsurface structure. For example a dipping resonator decreases the peak amplitude and increases the peak frequency of the spectrum. Effects of two-dimensional (2-D) subsurface structure (non-horizontal layering) have been successfully modelled and can be observed on the orthogonal Horizontal-to-Vertical Spectral Ratios, where N-S/V and E-W/V have different peak frequencies and peak amplitudes. By analysis of both horizontal spectral ratios the practitioner is able to determine whether one-dimensional (1-D) subsurface layering is present and hence 1-D assumptions are appropriate, e.g. inverting shear-wave velocities (Vs), calculating average Vs and depth to the resonator. HVSR measurements collected along seismic reflection profiles with known resonator topography, e.g. over steeply dipping bedrock resonators, are used to investigate both orthogonal horizontal components separately. Differences between the orthogonal H/V components are able to identify 2-D subsurface structure. Results demonstrate that extending analysis beyond the peak frequency to include orthogonal Horizontal-to- Vertical Spectral Ratios adds important information related to dipping surfaces and their orientations. Due to the ease and rapidity of HVSR data collection, the technique is ideally suited for reconnaissance scale survey work but also for infill where other data is sparse.