| Because the seismic exploration objects become more and more complex,accurate migration imaging of underground complex structures has received increasing attention.With the rapid development of computer performance in recent years,the reverse-time migration(RTM)technology has entered a stage of rapid development and is gradually applied in practical production.Compared with the acoustic case,the elastic RTM(ERTM)can provide abundant converted-wave and shear-wave imaging information.In practical processing,the imaging effects,amplitude preservation and computational efficiency of ERTM need to be further strengthened.This paper aims to improve the ERTM technique,and improve the practicability of ERTM.This paper mainly studies the two-way wave-equation-based ERTM imaging method and the source-independent converted-wave imaging method.The study focuses on amplitude preservation and computational efficiency of the elastic wavefield separation,the directional-wavefield separation imaging condition,analysis of imaging low-frequency noise,and other aspects.The main research results are as follows:(1)Five existing elastic wavefield separation methods are studied,and the amplitude preservation,computational efficiency and applicability in ERTM imaging are analyzed.Through combining least-squares-based optimal finite-difference method,an ERTM imaging method with amplitude-preserving wavefield separation is formed.In addition,to enhance the practicability of ERTM,it is necessary to take the applicability of different wavefield separation methods into overall consideration.(2)Considering that the wavefields used in cross-correlation imaging should conform to the kinematics and dynamics characteristics,especially at the imaging point,the ERTM imaging condition based on multi-directional-wavefield separation is developed.These operations can suppress low-frequency imaging noise and improve the quality of the imaging profile.Furthermore,to improve the calculation efficiency of the directional-wavefield separation,a one-step multi-directional P-and S-wavefield separation method with amplitude-preserving effect is constructed.The advantages of the ERTM conditions based on multi-directional-wavefield separation are verified by several modeling examples.(3)The low-frequency noise on the multi-directional-wavefield separation ERTM imaging profile is extracted and further used to identify the bottom interface of the near-surface low-velocity layer with the image recognition technology.Thus,it can obtain the interface information from the low-frequency imaging noise.The model test results illustrate the effectiveness of the method and provide a research direction for the application of machine learning technologies,such as image recognition,in traditional seismic exploration.(4)Considering the interference relationship between the P-wave and the converted S-wave during propagation,based on a pre-existing source-independent imaging condition,an improved source-independent converted-wave migration imaging condition is developed by combining with the directional-wavefield separation method.This converted-wave imaging condition can be used as a supplement and replacement when the ERTM imaging conditions are not applicable,and can further make full use of multi-component data to achieve imaging results.The feasibility is demonstrated by model tests. |
