Scholarlink

<p>Acquisition footprints are a pervasive form of coherent noise that degrades seismic imaging. This noise can compromise the relative amplitude fidelity required for amplitude-variation-with-offset (AVO) analysis. The effectiveness of transform-domain filtering may be limited when footprint noise aliases into the signal bandwidth. Additionally, model-based methods relying on global assumptions may risk smearing steeply dipping geological structures. To better balance noise suppression with structural preservation, this study proposes a novel Structure-Oriented Kriging (SOK) framework. The method begins by computing a stable local dip field via directional coherence scanning. We refine this field using Total Variation (TV) regularization. This step helps eliminate noise-induced oscillations while preserving sharp structural discontinuities. Subsequently, we incorporate this refined dip field to constrain the Kriging covariance model. By employing an anisotropic variogram and a dynamic search neighborhood aligned with local wavefronts, the estimator distinguishes stratigraphy from footprint artifacts based on spatial covariance. Synthetic tests on complex models indicate a notable improvement in the signal-to-noise ratio (SNR). Reflections are effectively recovered even in areas with conflicting dips. Application to 3D field data further demonstrates the removal of periodic striations without compromising fault or channel edge sharpness. The results suggest that SOK minimizes signal leakage and maintains amplitude integrity. This leads to reduced uncertainty in reservoir characterization and quantitative interpretation workflows.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Structure-Oriented Kriging for High-Fidelity Acquisition Footprint Suppression

  • 期刊论文

  • 发表日期：

    2026年2月19日

查看全文
  • Liang Huang,
  • Ming-ran Liu,
  • Wen-chao Su,
  • Jiang-yun Zhang,
  • Xu-hui Zhou,
  • Guo-xu Shu,
  • Shou-dong Huo

摘要

Acquisition footprints are a pervasive form of coherent noise that degrades seismic imaging. This noise can compromise the relative amplitude fidelity required for amplitude-variation-with-offset (AVO) analysis. The effectiveness of transform-domain filtering may be limited when footprint noise aliases into the signal bandwidth. Additionally, model-based methods relying on global assumptions may risk smearing steeply dipping geological structures. To better balance noise suppression with structural preservation, this study proposes a novel Structure-Oriented Kriging (SOK) framework. The method begins by computing a stable local dip field via directional coherence scanning. We refine this field using Total Variation (TV) regularization. This step helps eliminate noise-induced oscillations while preserving sharp structural discontinuities. Subsequently, we incorporate this refined dip field to constrain the Kriging covariance model. By employing an anisotropic variogram and a dynamic search neighborhood aligned with local wavefronts, the estimator distinguishes stratigraphy from footprint artifacts based on spatial covariance. Synthetic tests on complex models indicate a notable improvement in the signal-to-noise ratio (SNR). Reflections are effectively recovered even in areas with conflicting dips. Application to 3D field data further demonstrates the removal of periodic striations without compromising fault or channel edge sharpness. The results suggest that SOK minimizes signal leakage and maintains amplitude integrity. This leads to reduced uncertainty in reservoir characterization and quantitative interpretation workflows.