AUTOMATIC MICROSEISMIC EVENT LOCATION USING DEEP NEURAL NETWORKS IN ANISOTROPIC MEDIA

Accurate microseismic event location offers invaluable insights into the subsurface conditions not only for oil and gas production but also for seismic hazard assessment. Conventional microseismic event location methods face considerable drawbacks like requiring manual traveltime picking or large computational cost for simulating the wavefields. In fact, the need to locate microseismic events in real time leaves a gap for an automatic and efficient approach. Building on a previously developed method which is based on a deep Convolutional Neural Network for microseismic event location, we propose an extension of such an approach to include the anisotropic nature of the Earth and irregular receiver sampling. Example application on a 2D SEAM time-lapse model illustrates both the accuracy and efficiency of this method. Moreover, we validate the practicability of this approach for both isotropic and anisotropic media considering that the Earth is predominantly anisotropic. Equally important, we demonstrate that this approach is not only feasible for the cases with a uniform receiver distribution, but also applicable to cases where the passive seismic data are acquired with irregularly spaced receivers.