Research On Reverse Time Migration Method Of VSP Data Based On Traveling Wave Separation

With the continuous development of oil and gas exploration,the exploration target has changed from conventional reservoir to unconventional reservoir.Conventional ground seismic observation system has problems such as insufficient Vertical resolution and low signal-to-noise ratio.VSP(vertical seismic profiling)seismic data has the characteristics of high resolution and high signal-to-noise ratio.Due to the particularity of VSP observation system,the horizontal coverage of VSP data on underground structure is not uniform and the coverage times are low.The phenomenon of migration smile often exists in the reverse time migration(RTM)results,which will affect the interpretation accuracy,and even "submerge" the effective reflection wave in serious cases.In this paper,VSP data forward modeling,reverse time migration imaging,wave field separation,migration aperture prediction were studied.Firstly,the conventional grid format and the finite difference algorithm of rotated grid format were used to discretization the twodimensional constant-density acoustic wave equation,and the PML(Perfectly Matched Layer)was used to absorb the attenuation wave field boundary reflection;Secondly,in order to solve the problems of large memory consumption and low computational efficiency in the process of reverse time migration,the effective boundary conditions and random boundary conditions were applied to optimize the reverse time migration algorithm to reduce the memory consumption,and GPU acceleration technology was used to improve the calculation speed;Thirdly,for the low-frequency noise generated in the cross-correlation imaging process,the Hilbert transform method is used to separate the up and down wavefields,and the horizontal layered velocity model is established to verify the effectiveness of the wavefield separation method;Finally,two migration aperture prediction methods: Horizontal formation hypothesis method and Ray tracing method,were proposed.The actual migration velocity field was used to verify their suppression effect on migration smile phenomenon.The results show that both methods can effectively improve the migration smile phenomenon and improve the imaging accuracy of VSP data.The comprehensive analysis shows that the research content in this paper can further improve the reverse time migration imaging accuracy of VSP data,and the improvement of the reverse-time migration algorithm of VSP data has far-reaching research significance and broad application prospect for realizing the deep structure high-precision imaging around the well.