MODELING HVSR IN A LATERALLY INHOMOGENEOUS LAYERED MEDIUM USING THE DIFFUSE FIELD ASSUMPTION
The evaluation of the effects of surface geology is crucial for seismic hazard assessment. The horizontal-to-vertical spectral ratio (HVSR) of ambient seismic noise (ASN) is a very popular technique to assess the dominant soil frequencies of a given site. In combination with inversion schemes, it is very useful to estimate the properties of the subsoil assuming flat layers and usually ignoring lateral inhomogeneities. In practice, HVSR show significant lateral and azimuthal changes. This introduces uncertainties in the characterization of a site, therefore lateral and angular variations in assorted layered settings should be accounted for. Inversion of soil properties for horizontally layered media is feasible assuming that ASN constitutes a diffuse field, i.e., engendered by equipartitioned uniform illumination and/or by random sources and multiple scattering by heterogeneities. The HVSR have been modeled under the diffuse field assumption (DFA). This is achieved in practice by calculating the imaginary parts of the Green’s function (IMGs) when source and observer coincide at the same point. We use the 3D Indirect Boundary Element Method (IBEM) to model directional HVSR in a medium with lateral inhomogeneity. The IMGs at the source required to get HVSR has frequency dependent locality properties and may imply significant economies in the calculation. For simple models we modeled the IMGs approximately using an adaptive meshing scheme that accounts both for the locality of the problem and the diffraction properties of waves at low and high frequencies. The obtained directional HVSR detected variations in both its maximum value and the frequency at which it is maximum, for different azimuths. Layer interface variations can give raise to “spots” of large amplification. This suggest the need for denser field measurements for lateral and azimuthal variability. This work was partially supported by the UNAM-UI Research Partnership Program 2023 and by DGAPA UNAM under projects PAPIIT IN105523 and IN104823.