Receiver function arrays: a reflection seismic approach

The receiver function method (RFM) is a commonly used technique to study the crustal and upper mantle velocity structure. Early receiver function (RF) investigations were performed mostly at individual permanent stations. They were focused on crustal structures, and later on upper mantle velocity discontinuities (410 km and 660 km discontinuities). Only recently has research been directed towards the study of the lateral (2- and 3-D) variability of major velocity boundaries in the crust and upper mantle by receiver function arrays using temporary and permanent, three-component, short-period and broad-band seismic stations. To improve the signal-to-noise ratio, receiver functions are calculated for individual earthquakes and are then binned, moveout corrected and stacked. We show that this processing sequence is similar to that applied routinely in exploration seismology. Therefore, existing tools from the near-vertical data processing can be adopted for receiver functions: velocity analysis tools, solutions for static and residual static problems, coherence enhancement of seismic phases, migration, etc. The high spatial density of seismic stations of recent and future receiver function experiments provides the opportunity (and obligation) to use the more sophisticated migration methods (full wavefield migration) commonly and successfully used in exploration seismics. Synthetics calculated by the finite difference method for simple 2-D crustal models are employed here to test our processing approach and to show the potentials and limitations of stacking and migrating RF data. We show that binning, normal move-out (NMO) corrections, stacking and post-stack migration of the synthetic data can reconstruct the models reliably with a high spatial resolution.