Backprojection – Jing Ci Neo


Backprojection of Earthquakes

Earthquakes produce circular waves similar to ripples, which arrive at each station in a seismic array at a slightly different time. Back-projection (BP) is a method that locates the rupture of an earthquake by collapsing the wavefronts to find their centers. It can produce a “movie” of the earthquake rupture over time.

BP images often have artifacts (false peaks) and uncertainty, which may be due to (1) limited knowledge of the Earth’s velocity structure, (2) depth phases, (3) poor array coverage or (4) variations in focal mechanism etc. Many techniques have been used to reduce the artifacts and improve the accuracy of BP: cross-correlation, nth root stacking, phase-weighting, compressive sensing, and multiple-signal classification. However, results may still be uncertain if there are large near-source velocity heterogeneities or if the earthquake rupture is very complicated.

Animation of an earthquake rupture. Only a few waves are shown for simplicity, but in reality waves are continuously produced by the entire rupture

A New Method – Frequency-Difference Backprojection

We develop a new method, Frequency-Difference Backprojection (FDBP), to further improve the robustness of earthquake BP. The Frequency-Difference method was first developed in Naval Engineering to locate acoustic sources. It uses the “Autoproduct” of complex wavefields at two different frequencies to simulate a real wavefield at the difference frequency. As lower frequencies are less affected by velocity heterogeneities, FDBP has the potential to locate real source locations more accurately.

lsa logoum logoU-M Privacy StatementAccessibility at U-M