平面波最小二乘逆时偏移编码策略分析

最小二乘逆时偏移将成像视为最小二乘意义下的反演问题,相较于常规偏移方法具有更好的成像分辨率和保幅性,但其计算效率较低。为解决这一问题，推导了基于平面波编码的最小二乘逆时偏移方法,将海量炮数据压缩为若干个平面波记录,并发展了静态编码、动态编码、混合编码、随机动态编码等提高计算效率的编码策略。在优化算法基础上，对Marmousi模型进行了成像处理,并与常规逆时偏移进行了对比,结果表明：基于不同编码策略的平面波最小二乘偏移能够改善常规逆时偏移的成像质量和分辨率,并具有较高的计算效率;且不同的编码策略具有各自的优势,在对地震数据进行成像处理时,应根据需求选择合适的策略。

Least-square reverse time migration delivers high resolution and amplitude-preserved images as it introduces the imaging problem into the least-square inversion framework.However,it also has the defect of poor efficiency.The paper presents the theory and workflow of plane-wave least-square reverse time migration (PLSRTM) with different encoding strategies including statistic encoding,dynamic encoding,hybrid encoding and random dynamic encoding.Eefficiency improvement is achieved by encoding enormous shot data to produce several plane-wave records.Numerical tests on Marmousi model and comparison with RTM results show that PLSRTM can produce images with higher quality and resolution efficiently compared with conventional RTM; and appropriate strategies should be chosen according to the imaging requirements to implement the PLSRTM since every encoding strategy has their own advantages.

国家重点基础发展计划(973计划)项目(2014CB239006和2011CB202402)、国家自然科学基金项目(41104069和41274124)、中央高校科研业务费专项基金项目(R1401005A)及大地测量与地球动力学国家重点实验室开放基金项目(SKLGED2015-5-2-EZ)联合资助。