Generative Geostatistical Modeling from Incomplete Well and Imaged Seismic Observations with Diffusion Models

TL;DR

This paper presents a novel diffusion-based generative model for synthesizing subsurface velocity models from incomplete well and seismic data, enabling high-fidelity and uncertainty-aware subsurface modeling without extensive datasets.

Contribution

Introduces a diffusion generative model that synthesizes subsurface velocity models from incomplete data, overcoming the need for large fully sampled training datasets.

Findings

Accurately captures long-range geological structures.

Achieves high SSIM scores aligning with ground-truth models.

Provides meaningful uncertainty estimations for subsurface models.

Abstract

In this study, we introduce a novel approach to synthesizing subsurface velocity models using diffusion generative models. Conventional methods rely on extensive, high-quality datasets, which are often inaccessible in subsurface applications. Our method leverages incomplete well and seismic observations to produce high-fidelity velocity samples without requiring fully sampled training datasets. The results demonstrate that our generative model accurately captures long-range structures, aligns with ground-truth velocity models, achieves high Structural Similarity Index (SSIM) scores, and provides meaningful uncertainty estimations. This approach facilitates realistic subsurface velocity synthesis, offering valuable inputs for full-waveform inversion and enhancing seismic-based subsurface modeling.