Extracting surface wave dispersion curves from two-station microtremor analysis in heterogeneous ambient noise wavefield

The microtremor survey method (MSM) has been established as a method to estimate shallow S-wave velocity profiles from passive seismic data (i.e., microtremors). Although array analyses usually used in MSM are robust, a two-station microtremor analysis is attractive because it has a large potential to improve lateral resolution and applicability of MSM. However, applications of two-station microtremor analysis are limited probably because heterogeneous source distributions of microtremors expected in exploration studies (>1 Hz) are not considered in conventional processing. In this study, we propose a workflow to estimate surface-wave dispersion curves between two-stations considering heterogeneous distribution of microtremor energy for exploration data. In our approach, we estimate both phase and group velocity dispersion curves between each station pair, modeling azimuthal-dependent microtremor energy. Our field data application demonstrates that compared to conventional two-station analyses without consideration of noise heterogeneity, we can extract reliable phase velocity dispersion curves in wider frequency ranges. A number of dispersion curves between two-stations by our approach can be therefore useful to construct high-resolution 3D S-wave velocity structure by applying surface-wave tomography.