History Matching under Uncertainty of Geological Scenarios with Implicit Geological Realism Control with Generative Deep Learning and Graph Convolutions

2025-07-14

AI Summary

Addressing the challenge of simultaneously achieving high inversion accuracy and geological plausibility in history matching under geological uncertainty, this paper proposes an end-to-end framework based on a Graph-structured Variational Autoencoder (GVAE). The method explicitly models sedimentary and structural diversity while implicitly enforcing geological realism in the latent space via geodesic distance metrics. Integrating graph convolutional networks, principal component analysis (PCA), and topological data analysis (TDA), it overcomes the limitations of conventional grid-based deep learning in representing complex channelized geometries. Evaluated on 3D synthetic single- and double-channel datasets, the approach delivers high-fidelity geological reconstruction, robust uncertainty quantification, and efficient history matching. It significantly improves generalizability across diverse geological scenarios and enhances physical consistency with subsurface processes.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Web Mining and Content Analysis: Models for Web evolution

Abstract

The graph-based variational autoencoder represents an architecture that can handle the uncertainty of different geological scenarios, such as depositional or structural, through the concept of a lowerdimensional latent space. The main difference from recent studies is utilisation of a graph-based approach in reservoir modelling instead of the more traditional lattice-based deep learning methods. We provide a solution to implicitly control the geological realism through the latent variables of a generative model and Geodesic metrics. Our experiments of AHM with synthetic dataset that consists of 3D realisations of channelised geological representations with two distinct scenarios with one and two channels shows the viability of the approach. We offer in-depth analysis of the latent space using tools such as PCA, t-SNE, and TDA to illustrate its structure.

Problem

• Handling uncertainty in geological scenarios using generative deep learning
• Controlling geological realism implicitly with latent variables and metrics
• Analyzing latent space structure with PCA, t-SNE, and TDA tools

Innovation

• Graph-based variational autoencoder handles geological uncertainty
• Implicit geological realism control via latent variables
• Geodesic metrics ensure realistic geological representations