一种优化系数的混合域叠前深度偏移方法

A new prestack depth migration method in hybrid domain with optimization coefficient

为提高偏移中频散方程单平方根算子渐近式的逼近程度,提出了优化系数的混合域叠前深度偏移方法。该方法对频散方程的单平方根算子采用了有理切比雪夫逼近;与连分式展开的逼近算法对比后发现,该优化算法能降低偏移逼近算子与频散方程的单平方根算子的相对误差,从而提高了在陡倾构造及强横向速度变化地区偏移成像的精度。二维SEG/EAGE盐丘模型的偏移成像结果证明:在陡倾角构造及横向速度变化剧烈的地区,优化系数的混合域叠前深度偏移方法比常规的傅里叶有限差分(FFD)法的成像效果更好。

To improve the approximation degree of single square-root operator of frequency dispersion, prestack depth migra- tion by hybrid domain with optimization coefficient was proposed. This method is designed to calculate the optimization coef- ficient by rational Chebyshev approximation to approach the single square-root operator of frequency dispersion. By compari- son on the rational Chebyshev approximation and continued fraction approximation, it is discovered that the rational Cheby- shev approximation can reduce the relative error between optimization coefficient migration operator and single square-root operator of frequency dispersion. Therefore, our method can improve the precision of migration imaging in the area with high- steep structures and strong lateral velocity variation. The application on 2-D SEG/EAGE salt dome model proves the prestack depth migration in hybrid domain with optimization coefficient has better imaging effect than conventional Fourier finite difference （FFD） in the area with high-steep structures and strong lateral velocity variation.