arXiv subjects

Suihong Song

Publications and source records attributed to Suihong Song.

2 recordsLinked to original sources

arXiv2026

FMSIM: A Multimodal Flow Matching Framework for Conditional Geomodeling

Subsurface geomodeling plays a critical role in reservoir characterization, uncertainty quantification, and subsurface flow prediction. However, integrating heterogeneous sources of geological information, including conceptual geological descriptions, sparse well observations, and spatial prior constraints, remains a significant challenge for traditional geostatistical and data-driven geomodeling approaches. In this study, we present FMSIM, a multi-modal conditional flow matching framework for subsurface facies model generation. FMSIM utilizes a deep learning formulation to learn a velocity field that transports samples from a simple prior distribution to a complex geological facies distribution. Global geological semantic information is incorporated through a learned semantic representation framework and a learned prior model, while local hard constraints are enforced via an iterative projection strategy during sampling to ensure 100% fidelity to well observations. Additionally, a temporal guidance gating mechanism is introduced to regulate the influence of spatial probability maps, balancing large-scale trend alignment with fine-scale geological variability. Benefiting from the framework design, the model enables efficient and stable training with a simple loss function. The framework's fully convolutional architecture also demonstrates promising generalization to moderately larger grid sizes not seen during training without retraining. Results on a synthetic fluvial channel dataset indicate that FMSIM captures complex non-stationary geological features and produces geologically consistent realizations under multi-modal conditioning. This approach offers a flexible tool for incorporating conceptual geological knowledge, sparse observational data, and spatial priors into probabilistic subsurface geomodeling workflows.

physics.geo-ph↗

arXiv2026

DiffSIM: Unconditional and conditional facies simulation based on denoising diffusion generative models

Constructing subsurface facies models that are geologically plausible and constrained by well facies is essential for analyzing sedimentary evolution, reservoir characterization, and flow simulation. Recent deep generative model-based geomodelling methods have demonstrated promising capabilities for both unconditional and conditional settings. We investigate denoising diffusion models as a generative framework for producing realistic and diverse facies realizations in both settings. Diffusion models generate samples through a fixed forward noising process and a learned reverse denoising process. For unconditional geomodelling, we use denoising diffusion probabilistic models (DDPMs) to learn geological patterns from training facies models, and adopt denoising diffusion implicit models (DDIMs) to accelerate sampling by reducing inference steps from 1500 to 50 (30x fewer steps). We assess the geological plausibility using data distribution, class proportions, variograms, and geometric features. To enable conditional generation, we encode well facies and their spatial locations as conditional indicators and apply a mask-based denoising strategy that generates facies only in between-well regions, guaranteeing hard conditioning without introducing additional loss weights. We evaluate both unconditional and conditional generation on three scenarios: two two-dimensional cases (meandering channels and point bars) and one three-dimensional point-bar case. Across these scenarios, unconditional generation reproduces geological realism, and conditional generation honors well data while producing geologically consistent between-well realizations, demonstrating practical utility for facies geomodelling applications.

physics.geo-ph↗