Quantifying Diffracted and Passive Event Location Uncertainties with Local Wavefront Measurements

The quantification of uncertainty in diffraction or passive source depth localization is a major goal in current research. Stacking schemes like the common-reflection-surface stack are known to be reasonably reliable even if the data is strongly noise-contaminated, which is particularly useful in passive seismic settings, but also for diffractions, which are generally faint compared to reflections. As a by-product to the stack, these methods result in local wavefront attribute estimates, which can be used for efficient and automated joint velocity model building and localization. While most existing approaches to assess uncertainty require to perform many subsequent inversions, we suggest a simple and fully automatic strategy to globally link those independently performed local coherence measurements that share the same origin in depth, i.e. a common passive source or diffractor position. Through this global characterization of events, we show that it is possible to perform event-consistent statistics, which allow to estimate mean scattering or passive source locations and conveniently assess localization uncertainty. A synthetic diffraction example proved the feasibilty of the suggested scheme, in that it could successfully improve the localization and, in addition, provide useful insight into the convergence behavior of wavefront tomography, which might be useful for future applications.