基于弹性参数分解的横向各向同性介质弹性波波场分离

弹性波波场解耦是各向异性弹性波逆时偏移的关键环节之一。基于Helmholtz分解的波场解耦方法和基于时空域各向同性的波场解耦方法在各向异性介质中解耦不彻底, 解耦后纵横波存在串扰; 波数域各向异性波场解耦方法由于在每个时间节点需要进行傅里叶变换而导致计算量巨大。为此, 将横向各向同性(VTI)介质中弹性参数分解为纵波参数和横波参数, 分别将纵波参数和横波参数代入弹性波方程得到解耦的纵、横波方程, 从而在波场延拓的过程中实现纵、横波矢量分解。该方法计算量小, 无需进行振幅和相位校正, 有利于各向异性弹性波逆时偏移。采用简单模型对弹性波波场分离效果进行了测试, 发现分离结果中存在少量残余。对分解后的矢量波场抽取单道波形进行残差分析, 并利用标准模型进行弹性波逆时偏移成像测试, 成像结果反映界面信息一致且信噪比较高, 对复杂构造如断层、褶皱等的刻画清晰, 没有明显的成像假象以及深度不一致等现象, 测试结果表明了该方法的有效性。

The decomposition of elastic wavefields is key to anisotropic elastic reverse-time migration(ERTM).Because the P- and S-waves in anisotropic materials are not polarized parallel and perpendicular to the wave vectors, respectively, the wavefield decoupling method based on Helmholtz decomposition and that based on isotropy in the time-space domain are not suitable for anisotropic media.This unsuitability is because their use results in crosstalk of P- and S-waves after decoupling, and ultimately affects the imaging quality.The anisotropic wavefield decoupling methods in the wavenumber domain are computationally expensive owing to the Fourier transform at each period.In this study, we propose a time-space domain wavefield decomposition method, which first decomposes the elastic parameter in transversely isotropic(VTI) media into the P- and S-wave constants, and then substitutes them into the elastic wave equation to obtain the decomposed P- and S-elastic equations.According to this process, we can obtain the vector P- and S-waves.This method is simple and does not require amplitude or phase correction, which will promote the industrialization of elastic reverse-time migration in VTI media.A simple model was used to test the separation effect of the elastic wavefield, which revealed that little crosstalk still exists after this process.Moreover, a single trace was used to analyze the crosstalk of the decomposed vector wavefield.A standard model was used to test the elastic reverse-time migration.The imaging results show that the interface information is consistent, has a high signal-to-noise ratio, and can clearly describe complex structures, such as faults and folds.Additionally, no imaging artifacts or depth inconsistencies are present.The model test demonstrates the adaptability of the proposed method.