<p>In seismic waveform inversion, access to low-frequency information is crucial for an accurate reconstruction of the long-wavelength components of the subsurface model, but, unfortunately, seismic data often lack low-frequency information. One way to overcome this deficiency is to transform the seismic data into an envelope format that concentrates the spectral energy at lower frequencies compared to the original data. However, for salt models characterized by strong reflection contrasts and thick layers, the low-frequency content in conventional envelope data may not be low enough to effectively reconstruct the long-wavelength components from an initial linear model. To overcome this limitation, we adaptively select the window sizes for each trace, aiming to achieve much lower frequencies by smoothing the envelope. These window sizes are determined by analyzing the differences between the picked times corresponding to the upper and lower boundaries of impedance structure with large values, such as salt body, in the seismic profiles. With these dynamically selected trace-dependent window sizes, we implement a smoothed envelope waveform inversion procedure to improve the reconstruction of the long-wavelength components of the impedance model. This procedure is applied to three-dimensional seismic data where frequencies below 4&#xa0;Hz are missing. We then take this inverted impedance model as a starting point for a Fourier-based inversion, which further refines the impedance model. The tests performed on a 3D salt impedance model show good recovery of the salt body and provide evidence for the feasibility and effectiveness of the adaptive window size in the seismic waveform envelope inversion.</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

Seismic Waveform Inversion Using Smoothed Envelope Data for 3D Salt Impedance Model

  • Fengxia Gao,
  • Ying Rao,
  • Yanghua Wang

摘要

In seismic waveform inversion, access to low-frequency information is crucial for an accurate reconstruction of the long-wavelength components of the subsurface model, but, unfortunately, seismic data often lack low-frequency information. One way to overcome this deficiency is to transform the seismic data into an envelope format that concentrates the spectral energy at lower frequencies compared to the original data. However, for salt models characterized by strong reflection contrasts and thick layers, the low-frequency content in conventional envelope data may not be low enough to effectively reconstruct the long-wavelength components from an initial linear model. To overcome this limitation, we adaptively select the window sizes for each trace, aiming to achieve much lower frequencies by smoothing the envelope. These window sizes are determined by analyzing the differences between the picked times corresponding to the upper and lower boundaries of impedance structure with large values, such as salt body, in the seismic profiles. With these dynamically selected trace-dependent window sizes, we implement a smoothed envelope waveform inversion procedure to improve the reconstruction of the long-wavelength components of the impedance model. This procedure is applied to three-dimensional seismic data where frequencies below 4 Hz are missing. We then take this inverted impedance model as a starting point for a Fourier-based inversion, which further refines the impedance model. The tests performed on a 3D salt impedance model show good recovery of the salt body and provide evidence for the feasibility and effectiveness of the adaptive window size in the seismic waveform envelope inversion.