3D microseismic events localization using a 2D conditional diffusion model
D. Wamriew, T. Alkhalifah, M. Ravasi
Fifth International Meeting for Applied Geoscience & Energy (IMAGE 2025), SEG/AAPG, pp. 1462–1466 · 2025
Abstract
Microseismic event location in 3D can be an expensive and uncertain process, especially with limited irregular recording. We present a novel approach for 3D localization of microseismic events directly from passive seismic waveform gathers using a 2D conditional diffusion model. The influence of unknown origin time on the waveform gathers is mitigated by cross-correlating every trace from the gather by a reference trace from the same gather. The 3D microseismic event location coordinates (x, y, z) are projected to two 2D Gaussian heat maps in the x-y and x-z planes, enabling the use of a 2D conditional diffusion model for the inversion task in contrast to traditional localization techniques. The diffusion model is trained to generate accurate microseismic event locations conditioned on the waveform gathers. Tests on semi-synthetic TOC2ME field data show that this approach is computationally tractable while producing accurate 3D locations of the microseismic events. We will show the results of the application of this method on field data in the presentation of this work.
Related projects