Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
June 15, 2026Water Resources ResearchOpen Access

Cross‐Dimensional Generative Adversarial Networks (CDGAN) Geomodeling: Bridging 2D Geological Figures and 3D Reservoir Modeling

View Full Paper
Ask AI
Bookmark
Share

Authors

XHXun HuYYYanshu YinCZChangmin Zhang

Discussion

Loading...

Member takes

Overview

Randomized trial evaluates 3D reservoir generation from 2D geological figures, suggesting a novel data-driven approach for geological modeling.

Key Points

  • The aim is to develop a framework (CDGAN) that generates 3D geological models from limited 2D geological data.
  • Developed a cross-dimensional GAN framework integrating 3D generator and multiple 2D discriminators.
  • Utilized multi-discriminator strategy to enhance training robustness and stability.
  • Validated the method on synthetic and real-field cases, comparing results to existing GAN approaches.
  • CDGAN generated results closely resembling reference data with improved geological morphology compared to SliceGAN.
  • Conditional CDGAN achieved over 85% match rate with 1D well data; effectively constrained by low-resolution probability volumes.
  • Real applications showed that CDGAN can produce diverse 3D reservoir types consistent with geological patterns and sand distributions.

Cite This Study

Hu et al. (2026) studied this question.

synapsesocial.com/papers/6a2f97c8a1cfeec490828c63https://doi.org/10.1029/2025wr042708
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Data-Driven Development of Three-Dimensional Subsurface Geological Model from Limited Boreholes and Prior Geological Knowledge for Site Characterization2025
  2. 2A Physics-Constrained Two-Stage GAN for Reservoir Data Generation: Enhancing Predictive Accuracy2025
  3. 3Dynamic Reservoir Geological Mapping Using Generative AI and Satellite Data2025
  4. 4From simulator to surrogate: GANs for spatiotemporal modeling of subsurface flow in porous media2026
  5. 5High-Fidelity Seismic Image Generation From Limited Data Sets using Generative Adversarial Networks2025