3D Microseismic Image-Domain Elastic Velocity Inversion

Abstract

Summary Generating accurate and timely location estimates of microseismic events plays a crucial role in the success of seismic monitoring programs. Elastic time-reverse imaging offers a robust wavefield-based approach to obtain event location estimates, the quality of which is primarily dependent on the accuracy of the elastic velocity models. We present an elastic image-domain wavefield tomography method for passive data using zero-lag and extended source images. A multiterm objective function is designed to measure the focusing of imaged events in both zero-lag and extended domains. We apply numerous penalty operators to event images to isolate residual misfocusing energy emerged due to erroneous velocity models. An iterative minimization of the objective function leads to reliable P- and S-wave velocity model updates computed via the adjoint-state method. We present a realistic 3D synthetic numerical experiment to demonstrate that one can construct accurate elastic velocity models, which result in improved image focusing and a considerable reduction in event misposition error.