Calculation of Station-Representative Isotropic Receiver Functions

The estimation of one-dimensional (1-D) isotropic structures is routine work in most receiver function (RF) analyses that generally use a reference radial RF (RRF) for each station. However, the assumption of negligible back-azimuthal dependencies in a set of RRFs for a station may not be valid because of anisotropic layering, dipping structures, or incorrect sensor orientations. This work presents a comprehensive procedure to obtain a station-representative isotropic RRF, which can be applied automatically to prepare RRF data. The method incorporates a harmonic stripping method with a grid-search for sensor orientations. An optimum angle for the sensor orientation is determined by searching for the angle that minimizes the energy in the tangential component RF (TRF). For each searched angle, possible effects by anisotropy and dipping structures are iteratively suppressed by an inversion process to exclude two- and four-robe back-azimuthal patterns. The performance of the method was first confirmed with a test using a set of highly noisy composite RRFs and TRFs. The method was then applied to RF data from the southern Korean Peninsula and southwestern Japan. Obtained isotropic RRFs and measured station orientations were found to be reliable in comparisons with results from neighboring stations and previous studies. As an automatized routine pre-process, the obtained isotropic RRF data are particularly useful for estimating 1-D isotropic structures in migration or inversion studies, which are potentially affected by back-azimuthal dependencies in RF data calculated through conventional averaging methods.