基于照明补偿的变“子波”模拟成像方法

Variable“wavelet”simulated imaging method based on illumination compensation

叠前深度偏移是目前高精度地震成像的关键技术。对于给定的速度模型,传统方法是先利用波动方程做波场正演,再对正演波场进行叠前偏移成像。但该方法计算效率低,对硬件要求高。基于成像算子的模拟叠前深度偏移方法无需波场正演和偏移处理,计算效率高,但存在能量分布与实际不符、分区重构分界严重及成像波形不正确等问题。为改善模拟叠前偏移成像算法的效果,文中提出基于照明补偿的变“子波”模拟叠前深度偏移成像方法。首先根据观测系统计算射线路径,使用单位脉冲代替点扩散函数在波数域构建直接成像算子,并将成像算子作用于反射系数模型;然后在空间域通过非稳态褶积运算得到模拟深度域成像结果,解决模拟深度成像时波形不能空变的问题;通过单位脉冲作用下多点模拟成像结果的叠加,空间能量差异及个别区域无法照明的问题得以解决。该方法不仅保留了原算法计算速度快的优势,同时还解决了能量分配和“子波”空变的问题。模型试算证明该方法成像结果准确,可为观测系统选择和地震成像提供参考。

Currently prestack depth migration is the key technology of high‑precision seismic imaging.For a given velocity model,the traditional method is to solve the wave equation for wavefield forward modeling first,and then do prestack migration imaging for the forward wavefield.This method features low computational efficiency and high requirements for hardware.At present,the simulated prestack depth migration method based on an imaging operator does not need wavefield forward modeling and migration processing with high computational efficiency.However,there are some problems,such as inconsistency of the energy distribution with the actual situation,serious zonal reconstruction division,and incorrect imaging waveform.To improve the effect of simulated prestack migration imaging algorithms,a variable“wavelet”simulated prestack depth migration imaging method based on illumination compensation is proposed in this paper.This method first calculates the ray path according to the acquisition geometry,adopts the unit impulse instead of the point spread function to construct the direct imaging operator in the wavenum‑ber domain,and acts as the imaging operator on the reflection coefficient model.Then,the imaging results in the simulated depth domain are obtained by unsteady convolution in the spatial domain,which solves the problem that the waveform cannot change spatially in the simulated depth imaging.Through the superposition of multi‑point simulation imaging results under the action of unit impulse,the problems of spatial energy difference and illumination inability in individual areas are addressed.This method not only retains the advantage of the fast calculation speed of the original algorithm but also tackles the problems of energy distribution and“wavelet”space variation.Model test proves that the imaging results of this method are accurate,which can provide more accurate references for the selection of acquisition geometry and seismic imaging.