论文部分内容阅读
Elastic migration has been widely paid attention by employing the vector processing of multicomponent seismic data.Ray based elastic Kirchhoff migration has such properties as high flexibility and high efficiency.However,it has failed to solve many problems caused by multipath.On the other hand,elastic reverse-time migration(RTM)based on the two-way wave equation is known to be capable of dealing with these problems,but it is extremely expensive when applied in 3D cases and velocity model building.Based on the elastic Kirchhoff-Helmholtz integral,we calculate decoupled backward-continued wavefields by introducing elastic Green functions for P-and S-waves,which is expressed by a summation of elastodynamic Gaussian beams.The PP and polarity-corrected PS images are obtained by calculating the correlation between downward and decoupled backward-continued vector wavefields,where polarity correction is performed by analyzing the relation between the polarization direction of converted PS waves and incident angle on the interface.To a large extent,our method combines the high efficiency of ray-based migration with the high accuracy of wave-equation based reverse-time migration.Application of this method to multicomponent synthetic datasets from the fault model and Marmousi 2 model demonstrates the validity,flexibility and accuracy of the new method.
Elastic migration has been widely paid attention by employing the vector processing of multicomponent seismic data. Ray based elastic Kirchhoff migration has such properties as high flexibility and high efficiency. However, it has failed to solve many problems caused by multipath. On the other hand, elastic reverse-time migration (RTM) based on the two-way wave equation is known to be capable of dealing with these problems, but it is extremely expensive when applied in 3D cases and velocity model building. Based on the elastic Kirchhoff-Helmholtz integral , we calculate decoupled backward-continued wavefields by introducing elastic Green functions for P-and S-waves, which is expressed by a summation of elastodynamic Gaussian beams. The PP and polarity-corrected PS images are obtained by calculating the correlation between downward and decoupled backward-continued vector wavefields, where polarity correction is performed by analyzing the relation between the polarization direction of converted PS wav es and incident angle on the interface. To a large extent, our method combines the high efficiency of ray-based migration with the high accuracy of wave-equation based reverse-time migration. Application of this method to multicomponent synthetic datasets from the fault model and Marmousi 2 model demonstrates the validity, flexibility and accuracy of the new method.