论文详情
Quantitative and accurate description of oil and gas reservoirs is the goal of seismic exploration.In view of that,seismic imaging has advanced from locating reflection/scattering point to estimating angle reflection coefficients.In general,reverse time migration (RTM) is the most accurate method for imaging complex structure,and least squares migration (LSM) is the best choice for estimating angle reflection coefficients.LSM is based on the linearized inversion theoretical framework.In theory,LSM can remove the influence of poor illumination and balance amplitude,and can improve image resolution.However,these theoretical advantages have not been realized in practical applications and the application of LSM is still in experimental exploration.This paper briefly introduces the theory and advantages,then compares the application effects of data-domain iterative inversion and the image-domain non-iterative inversion algorithm.Some of the observed major challenges of LSM in practical application were as follows: the background velocity cannot meet the requirement of linearized inversion,Born modeling operator cannot simulate primary reflected/scattered wave in observed data well,noise does not satisfy the Gaussian distribution assumption,the wavelet is unknown,and the cost of computation is high.In short term,the application of LSM depends on reasonable data matching techniques,an appropriate regularization technique,and the approximation of the inverse Hessian matrix.Meanwhile,integrating LSM into full waveform inversion (FWI) is essential in the long term.
国家科技重大专项(2016ZX05014-001-002)和国家重点研发计划(2017YFC0602804-02)共同资助。