Enhancing small-scale target inversion in acoustic full-waveform inversion using an embedded-scattering-point staining algorithm

Full-waveform inversion (FWI) recovers high-resolution velocity models by iteratively minimizing the misfit between observed and modeled data, but its gradient is computed in the same way as reverse-time migration (RTM) and therefore inherits RTM’s sensitivity to illumination. We introduce a staining-enhanced acoustic FWI scheme that isolates the gradient contribution of a marked small-scale target using the embedded-scattering-point staining algorithm: within the marked region, the gradient correlation is computed with a staining wavefield excited only by energy that has interacted with the target, rather than the full wavefield, yielding a locally modified search direction without altering the gradient elsewhere. The gradient and step length are computed via the adjoint-state method and a three-point parabolic line search within a conjugate-gradient-based FWI framework, with the staining correction activated once conventional inversion has reduced the relative misfit below a preset threshold. On a simple three-layer model, staining-enhanced FWI recovers a small square anomaly with markedly higher structural similarity to the true model, as quantified by the structural similarity index, than conventional FWI, and removes a bow-tie-shaped residual-diffraction artifact that otherwise surrounds the anomaly. On a sub-model extracted and modified from the geologically realistic Sigsbee salt-body benchmark, staining-enhanced FWI recovers marked scatterer targets with sharp velocity spikes that conventional FWI fails to resolve, while both methods recover the surrounding structure comparably. We further show that the method tolerates a moderate mislocation of the marked target—up to a capture radius comparable to the target’s own spatial extent—beyond which the stained gradient instead focuses spuriously at the marked (incorrect) location, clarifying the method’s practical requirements and failure mode. Additional tests with a masked-conventional baseline, a null test without the anomaly, mislocated target regions, and noisy data (signal-to-noise ratio = 5 dB) are used to assess the role of the target mask and the method’s robustness.

Authors

Publication Details

Journal
Journal of seismic exploration
Published
2026-11-16
DOI
https://doi.org/10.36922/jse026350167
Primary Topic
Seismic Imaging and Inversion Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Enhancing small-scale target inversion in acoustic full-waveform inversion using an embedded-scattering-point staining algorithm

Yingming Qu, Zhenchun Li, Jiajun Song, Chunbo Ma
Journal of seismic exploration
Seismic Imaging and Inversion Techniques
article

Enhancing small-scale target inversion in acoustic full-waveform inversion using an embedded-scattering-point staining algorithm

Yingming Qu, Zhenchun Li, Jiajun Song, Chunbo Ma
article en

Abstract

Full-waveform inversion (FWI) recovers high-resolution velocity models by iteratively minimizing the misfit between observed and modeled data, but its gradient is computed in the same way as reverse-time migration (RTM) and therefore inherits RTM’s sensitivity to illumination. We introduce a staining-enhanced acoustic FWI scheme that isolates the gradient contribution of a marked small-scale target using the embedded-scattering-point staining algorithm: within the marked region, the gradient correlation is computed with a staining wavefield excited only by energy that has interacted with the target, rather than the full wavefield, yielding a locally modified search direction without altering the gradient elsewhere. The gradient and step length are computed via the adjoint-state method and a three-point parabolic line search within a conjugate-gradient-based FWI framework, with the staining correction activated once conventional inversion has reduced the relative misfit below a preset threshold. On a simple three-layer model, staining-enhanced FWI recovers a small square anomaly with markedly higher structural similarity to the true model, as quantified by the structural similarity index, than conventional FWI, and removes a bow-tie-shaped residual-diffraction artifact that otherwise surrounds the anomaly. On a sub-model extracted and modified from the geologically realistic Sigsbee salt-body benchmark, staining-enhanced FWI recovers marked scatterer targets with sharp velocity spikes that conventional FWI fails to resolve, while both methods recover the surrounding structure comparably. We further show that the method tolerates a moderate mislocation of the marked target—up to a capture radius comparable to the target’s own spatial extent—beyond which the stained gradient instead focuses spuriously at the marked (incorrect) location, clarifying the method’s practical requirements and failure mode. Additional tests with a masked-conventional baseline, a null test without the anomaly, mislocated target regions, and noisy data (signal-to-noise ratio = 5 dB) are used to assess the role of the target mask and the method’s robustness.

Journal of seismic explorationVol. 0(0)
Seismic Imaging and Inversion Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon