Shallow-to-deep velocity model building via diffusion models-Part II: Realistic scenarios
Abstract
Full-waveform inversion (FWI) requires accurate initial velocity models to avoid cycle-skipping, but constructing such models remains challenging in practice. Building on the depth-progressive diffusion framework introduced in Part~I, which relied on idealized reflectivity constraints, this work adapts the methodology to realistic exploration scenarios. We replace perfect structural information with migration-derived attributes extracted from seismic images, and introduce smooth background velocity models from tomography as additional conditioning inputs. The framework jointly leverages background/migration velocity, migrated structural information, and sparse well measurements to synthesize high-resolution velocity models through depth-progressive generation. Validation on synthetic examples demonstrates superior accuracy compared to conventional interpolation and alternative deep learning methods, with generated models successfully initializing FWI and mitigating cycle-skipping even in complex geological structures. Field data confirms practical applicability: despite training on synthetic data, the method generalizes effectively to field conditions, producing velocity models with synthetic data response that nearly match observed seismic data. As a result, this framework establishes a practical pathway to deploy generative diffusion models for velocity model building under realistic constraints.
Related
- Shallow-to-deep velocity model building via diffusion models-Part I: Method and Proof of concept
- Depth-progressive 3D seismic velocity model building via 2D generative diffusion models
- Skillful and realistic precipitation forecasting with reinforced diffusion
- Self-attention summary networks for subsurface velocity-model building from common-image gathers
- Bayesian Joint Velocity and Impedance Inversion via Diffusion Models Conditioned on Common Image Gathers
BibTeX
@misc{cheng2026shallowtodeep,
title = {{Shallow-to-deep velocity model building via diffusion models-Part II: Realistic scenarios}},
author = {Shijun Cheng and Randy Harsuko and Tariq Alkhalifah},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.26482},
eprint = {2609.26482},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.26482},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Shallow-to-deep velocity model building via diffusion models-Part II: Realistic scenarios
AU - Shijun Cheng
AU - Randy Harsuko
AU - Tariq Alkhalifah
PY - 2026
DA - 2026/09/22
PB - arXiv
DO - 10.48550/arxiv.2609.26482
UR - https://arxiv.org/abs/2609.26482
AB - Full-waveform inversion (FWI) requires accurate initial velocity models to avoid cycle-skipping, but constructing such models remains challenging in practice. Building on the depth-progressive diffusion framework introduced in Part~I, which relied on idealized reflectivity constraints, this work adapts the methodology to realistic exploration scenarios. We replace perfect structural information with migration-derived attributes extracted from seismic images, and introduce smooth background velocity models from tomography as additional conditioning inputs. The framework jointly leverages background/migration velocity, migrated structural information, and sparse well measurements to synthesize high-resolution velocity models through depth-progressive generation. Validation on synthetic examples demonstrates superior accuracy compared to conventional interpolation and alternative deep learning methods, with generated models successfully initializing FWI and mitigating cycle-skipping even in complex geological structures. Field data confirms practical applicability: despite training on synthetic data, the method generalizes effectively to field conditions, producing velocity models with synthetic data response that nearly match observed seismic data. As a result, this framework establishes a practical pathway to deploy generative diffusion models for velocity model building under realistic constraints.
ER -