摘要
与共炮高斯波束偏移相比,共炮检距高斯波束偏移具有直接抽取炮检距域共成像点道集的优势.过去,共炮检距高斯波束偏移以损失成像精度的代价采用最速下降法来降低积分的维数,从而提高计算效率.但经过最速下降近似简化的偏移公式仍是频率域的,需要在每个频点进行计算.为此,本文提出一种快速实现算法来避免采用最速下降法.本文通过分析一个水平层状速度模型的偏移过程和Marmousi速度模型的成像结果来检验不同插值方法对快速实现算法的成像精度和计算效率的影响,并建议采用二维三次卷积插值方法.同时本文在Marmousi速度模型下验证了快速实现算法相对于最速下降法在成像精度和计算效率上的优势.此外,本文将采用二维三次卷积插值的快速实现算法应用于Sigsbee2A模型并获得了清晰的盐下图像.
Compared with the common-shot Gaussian beam migration, the common-offset Gaussian beam migration has the advantage in direct extraction of offset-domain common-image gathers. In the past, the common offset Gaussian beam migration adopted the steepest descent method to reduce the dimension of the integrals at the cost of losing precision and to increase computation efficiency. However, the migration formula simplified by the steepest descent method is still in the frequency domain, and it has to be evaluated at each frequency. To solve this problem, we develop a fast algorithm to avoid adopting the steepest descent method. We test the effects of different interpolation methods on the imaging accuracy and the computational efficiency of the fast algorithm through analyzing the migration procedures on a horizontally layered velocity model and the imaging results of the Marmousi velocity model. We suggest to adopt the bicubic convolution interpolation method. The superiority of this fast algorithm on the imaging accuracy and the computational efficiency compared with the steepest descent method is validated with the Marmousi velocity model. Furthermore, we apply this fast algorithm using the bicubic convolution interpolation to the Sigsbee2A model and obtain clear subsalt image.
出处
《地球物理学报》
SCIE
EI
CAS
CSCD
北大核心
2017年第3期1118-1131,共14页
Chinese Journal of Geophysics
基金
国家自然科学基金(41274120
41404085
41504084)
国家科技专项(Sinoprobe09-01)联合资助
关键词
共炮检距
高斯波束偏移
快速实现算法
插值方法
Common-offset
Gaussian beam migration
Fast algorithm
Interpolation method
