Nov 2022
Seeing the geology of the subsurface with accuracy and resolution remains a big challenge for seismic imaging. GeoMind is an innovative approach that integrates geology and geophysics into a scheme that generates unlimited training datasets that will enhance the prediction accuracy with time. Geologists use their priory information about an area, wells, and outcrops, to build high-resolution 3D geological models. Deep learning then generates nonstop scenarios of such models. We use elastic modeling program to simulate the seismic response of each model. This continuous process that runs on one of the top 10 supercomputers in the world keep enhancing the prediction accuracy of GeoMind. Implementations of this method on synthetic and real data show a promising approach to better image the subsurface.
Saleh Al-Saleh is heading the Geophysical Imaging Department in Saudi Aramco. He also led several teams such as Red Sea Exploration, Unconventional Gas, Depth Imaging, and Digital Transformation. He received his BSc from the University of Pacific in California, and MSc from University of Texas at Austin, and PhD from the University of Calgary. Saleh completed a Visiting Scholar Program with Stanford University and the Advanced Management Program with Wharton Business School. Saleh co-chaired several international workshops on topics related to exploration and development. He authored several technical papers with several best paper and honorable mention awards.