Scholarlink

<p>Lithological and physical property variations in subsurface sand bodies lead to highly complex seismic phase information. Accurately extracting, analyzing, and utilizing phase information from seismic data remains one of the challenging issues in identifying and predicting subsurface reservoirs using seismic data. Previous phase domain reconstruction methods rely on single-trace time–frequency analysis and neglect the time-varying characteristics of seismic wavelets during propagation, resulting in insufficient accuracy for identifying reservoir boundaries and special geological bodies. To address this issue, this paper proposes a phase domain reconstruction method driven by time-varying wavelet deconvolution: First, high-precision time–frequency analysis is achieved using the generalized S-transform, and the time-varying wavelet spectrum is extracted by optimizing time–frequency resolution parameters. Second, phase shifting over <i>θ</i> ∈ [ − <i>π</i>, <i>π</i>] is applied to the zero-phase time-varying wavelet to generate an arbitrary-phase wavelet <i>w</i>(<i>t</i>, <i>θ</i>). Finally, the reflection coefficient <i>r</i>(<i>t</i>, <i>θ</i>) is obtained by deconvolving <i>w</i>(<i>t</i>, <i>θ</i>) with the original seismic data. Convolution is then performed to generate specific phase component data <i>s</i>(<i>t</i>, <i>θ</i>), which are combined into phase gathers. The breakthrough of this method lies in overcoming the limitation of conventional methods that assume wavelet stationarity. In practical application, it enables multi-phase collaborative characterization of concealed channel sand bodies, effectively separates reservoir response characteristics, eliminates mixed phase interference, and provides a novel approach for the multi-scale characterization of tight sandstone reservoirs.Query</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

Research on phase domain seismic data reconstruction based on time-varying wavelet deconvolution

  • 期刊论文

  • 发表日期：

    2026年2月13日

查看全文
  • Wei Yang,
  • Junjie Wang,
  • Qinggao Zeng,
  • Linke Song,
  • Tao Yang,
  • Tanglv Li

摘要

Lithological and physical property variations in subsurface sand bodies lead to highly complex seismic phase information. Accurately extracting, analyzing, and utilizing phase information from seismic data remains one of the challenging issues in identifying and predicting subsurface reservoirs using seismic data. Previous phase domain reconstruction methods rely on single-trace time–frequency analysis and neglect the time-varying characteristics of seismic wavelets during propagation, resulting in insufficient accuracy for identifying reservoir boundaries and special geological bodies. To address this issue, this paper proposes a phase domain reconstruction method driven by time-varying wavelet deconvolution: First, high-precision time–frequency analysis is achieved using the generalized S-transform, and the time-varying wavelet spectrum is extracted by optimizing time–frequency resolution parameters. Second, phase shifting over θ ∈ [ − π, π] is applied to the zero-phase time-varying wavelet to generate an arbitrary-phase wavelet w(t, θ). Finally, the reflection coefficient r(t, θ) is obtained by deconvolving w(t, θ) with the original seismic data. Convolution is then performed to generate specific phase component data s(t, θ), which are combined into phase gathers. The breakthrough of this method lies in overcoming the limitation of conventional methods that assume wavelet stationarity. In practical application, it enables multi-phase collaborative characterization of concealed channel sand bodies, effectively separates reservoir response characteristics, eliminates mixed phase interference, and provides a novel approach for the multi-scale characterization of tight sandstone reservoirs.Query