LAUSR

Internal multiples bring great challenges for conventional seismic data imaging and interpretation, in which primaries only are regarded as signals. It is of great importance to suppress the internal multiples before conventional imaging, which is computationally expensive. Considering that internal multiples are also real reflections from the underground interfaces and reverse time migration (RTM) can realize the imaging of multiples. Instead of suppression before imaging, we propose to do the joint imaging of primaries and internal multiples based on RTM. However, the artifacts caused by the internal multiples should be suppressed to guarantee the precision. First, the imaging condition based on up-going and down-going extrapolated wavefield separation is introduced to improve the imaging accuracy. Then, the generation mechanism of image artifacts is analyzed, and it concludes that nonphysical wave paths caused by the missing boundary wavefields under the surface result in the infidelity reversed-time backward-extrapolation, which is the main reason for the artifacts. Next, we propose to apply the modeling boundary wavefields with a match based on pseudomultichannel matching filter to compensate the missing boundary wavefields, which can avoid the image artifacts and realize high-precision joint imaging. Finally, the feasibility and effectiveness of the proposed method are verified by numerical examples and field data.