基于构造滤波的VTI介质平面波最小二乘FFD叠前深度偏移

中国中东部地区沉积地层广泛发育，地下介质中地震波传播速度不仅与空间位置有关，还受传播方向影响，研发适用于各向异性介质的成像方法有利于提高反演成像精度。通过将VTI介质高阶傅里叶有限差分波场延拓算子纳入反演成像体系，本文实现了基于VTI介质的最小二乘FFD叠前深度偏移。针对该方法计算量大、相干成像噪声明显的不足，进行了优化：采用引入平面波编码策略，压缩地震数据，显著提高计算效率；迭代过程中提取CIG道集，在该道集上应用平面波构造滤波算子压制相干成像噪声，提高成像信噪比。基于各向异性Marmousi 2模型的数值测试结果表明，本文的优化偏移成像方法取得三点改进： ①适用于复杂强横向变速VTI介质； ②成像信噪比提高，效率更高； ③克服了单程波类算法近地表成像精度低的缺陷。

Structure-constrained plane-wave least-squares FFD prestack depth migration for VTI media

For most of the sedimentary strata in the central and eastern China,seismic wave velocity is determined by both the underground position and the spread direction.It is important to develop migration methods for anisotropic media to improve seismic imaging.With the high order Fourier finite difference (FFD) operator to the linear inversion imaging,we propose a VTI least-squares FFD prestack depth migration method (VTI-LSFFD).Based on the accomplishment of VTI-LSFFD,optimization strategies are adopted to solve some key problems of the least-squares migration such as heavy computation cost and sensitivity to coherent imaging noises.First,the plane-wave encoding strategy is used to reduce the amount of seismic data and improve computation efficiency.Then common imaging gathers (CIG) are extracted and plane-wave construction (PWC) operators are applied to the CIGs,the coherent imaging noise is suppressed and the signal-to-noise ratio (SNR) is improved.Based on our numerical experiments on the modified anisotropy Marmousi 2 model,the following advantages of the proposed method are found:this method is suitable for strong horizontal velocity variation; it has both high computation efficiency and low coherent imaging noises sensitivity; and with the plane-wave encoding strategy,it achieves better near-surface imaging than the one-way wave migration methods.