Abstract:

Reconstructing 3D structural geology from sparse surface and borehole observations is a longstanding challenge with critical applications in mineral exploration, geohazard assessment, and geotechnical engineering. This inherently ill-posed problem is often addressed by classical geophysical inversion methods, which typically yield a single maximum-likelihood model that fails to capture the full range of plausible geology. In this talk, I present StructuralGeo, a rapid simulation engine that generates synthetic lithological models encoding geological knowledge such as the tectonic, magmatic, and sedimentary processes found in textbooks. Using this engine as a synthetic dataset, we train both unconditional and conditional generative AI models that can reconstruct multiple plausible 3D scenarios from surface topography and sparse borehole data, depicting structures such as layers, faults, folds, and dikes. 

Bio:

Simon Ghyselincks is a researcher at the Vector Institute for Artificial Intelligence and a graduate student in the Department of Computer Science at the University of British Columbia (UBC). Co-advised by Professors Eldad Haber and Evan Shelhamer, his research focuses on the intersection of deep learning and geophysical modeling, specifically developing generative AI models for probabilistic solutions to complex inverse problems. Simon is a recent alumnus of the UBC Engineering Physics program (BASc) and a member of the UBC Institute of Applied Mathematics (IAM). In addition to his academic research, he works as a software consultant for Earth Dynamics AI, where he bridges the gap between machine learning research and practical geological applications. His work on synthetic geology and structural modeling was recently published in the Journal of Geophysical Research.