How Much Can We Hope to Resolve in Earthquake Rupture Processes with Back-projection

Abstract:

Back-projection has been proven to be reliable and effective for unraveling complicated earthquake rupture processes. It is very robust because the method makes few a priori assumptions about the fault geometry or rupture speed, and is relatively insensitive to 3D velocity variations. As most studies use array data at high frequencies for back-projection imaging, the results sometime suffer from artifacts, limited resolution, and unclear physical explanations. We have found that improved back-projection results can be obtained utilizing global data. First, global data often can provide fairly uniform azimuthal coverage, which improves spatial resolution and reduces back-projection artifacts, permitting hidden features in the ruptures to be studied in detail. Second, the good azimuthal coverage also enables back-projection to be performed at relatively low frequencies (0.05 to 0.2 Hz), which can fill in the gap between moment tensor/finite-fault inversions (low frequency content) and high-frequency back-projection/beam-forming. Third, P-wave polarity differences among global stations will affect the maximum coherency of back-projected power as a function of source location, which can be used to resolve the spatially varying focal-mechanisms of complicated earthquakes involving multiple fault segments. We plan to conduct synthetic tests to explore the resolution and uncertainty limits of global back-projection across multiple frequency bands. Ultimately, we hope to extend the potential of back-projection methods with global array data, while exploring the theoretical limits of the method.