基于点扩散函数的深度域褶积正演方法

打开文本图片集

中图分类号：P631 文献标识码：A DOI：10.13810/j.cnki.issn.1000-7210.20240287

Abstract：Seismic wave impedance inversion is thecore of reservoir prediction，and the construction ofanacurate seismic forward model is the basis of realizing high-resolution wave impedance inversion. However， it is difficult for the traditional convolution forward models based onone-dimensionalseismic wavelet to accurately simulate and characterize the imaging profile in the depth domain when there is a drastic lateral change in the subsurface velocity，which seriously affects the accuracy and reliability of wave impedance inversion results. For this reason，this paper proposes a forward modeling method of convolution in the depth domain of non-stationary space based on point spread function. Firstly，the accurate mapping relationship between seismic imaging profile andunderground reflection coefficient is derived according to the linear Born forward modeling theory.Then，an accurate convolution forward model is constructed with the point spread function as the seis mic wavelet in the multi-dimensional depth domain.Finaly，the efficient algorithm of point spread function based on Green's function of the ray theory can greatly improve the computational eficiency of the point spread function. The correctness and efectiveness ofthe proposed method are verified by the depth-domain convolution forward modeling tests of the simple horizontal layered model and the complex Marmousi model. Keywords： depth domain，wave impedance inversion，point spread function，convolutional model

沈蔚，岳玉波，秦宁，[J].石油地球物理勘探，2025，60（4）：951-957. SHENWei，YUE Yubo，QINNing：Depth-domain convolution forward modeling methodbased on point pread function[J]. Oil Geophysical Prospecting，2025，60（4） ：951-957.

0 引言

众所周知，地震正演模型是地震反演方法的基础[1]。（剩余8636字）