Generative Latent Diffusion Model for Inverse Modeling and Uncertainty Analysis in Geological Carbon Sequestration

2025-08-17

AI Summary

High uncertainty in subsurface flow modeling for geological carbon sequestration (GCS) arises from sparse observations and reservoir heterogeneity, while conventional inverse modeling and uncertainty quantification suffer from prohibitive computational cost and poor generalizability. Method: We propose the Conditional Neural Field Latent Diffusion Model (CNF-LDM), the first framework integrating conditional neural field representation with Bayesian latent-space diffusion to enable zero-shot cross-task generation, Bayesian posterior sampling, and data assimilation—without retraining. It employs self-supervised pretraining and latent-space data assimilation to establish an end-to-end uncertainty-aware modeling paradigm. Results: Validated on synthetic and real-world GCS scenarios, CNF-LDM achieves over 100× speedup versus traditional MCMC, significantly improves generalization and robustness, and enables high-fidelity forward/inverse simulation and uncertainty quantification under complex geometries—establishing a novel paradigm for intelligent geoscience-based energy decision-making.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Relational Probabilistic Models
• Computer Vision: Diffusion Models for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Geological Carbon Sequestration (GCS) has emerged as a promising strategy for mitigating global warming, yet its effectiveness heavily depends on accurately characterizing subsurface flow dynamics. The inherent geological uncertainty, stemming from limited observations and reservoir heterogeneity, poses significant challenges to predictive modeling. Existing methods for inverse modeling and uncertainty quantification are computationally intensive and lack generalizability, restricting their practical utility. Here, we introduce a Conditional Neural Field Latent Diffusion (CoNFiLD-geo) model, a generative framework for efficient and uncertainty-aware forward and inverse modeling of GCS processes. CoNFiLD-geo synergistically combines conditional neural field encoding with Bayesian conditional latent-space diffusion models, enabling zero-shot conditional generation of geomodels and reservoir responses across complex geometries and grid structures. The model is pretrained unconditionally in a self-supervised manner, followed by a Bayesian posterior sampling process, allowing for data assimilation for unseen/unobserved states without task-specific retraining. Comprehensive validation across synthetic and real-world GCS scenarios demonstrates CoNFiLD-geo's superior efficiency, generalization, scalability, and robustness. By enabling effective data assimilation, uncertainty quantification, and reliable forward modeling, CoNFiLD-geo significantly advances intelligent decision-making in geo-energy systems, supporting the transition toward a sustainable, net-zero carbon future.

Problem

• Addressing subsurface flow uncertainty in carbon sequestration
• Overcoming computational intensity in inverse geological modeling
• Enabling zero-shot generation for unobserved reservoir states

Innovation

• Combines neural field encoding with diffusion models
• Enables zero-shot conditional generation of geomodels
• Allows data assimilation without retraining