Surface-related Multiple Elimination Through Iterative Inversion

In seismic data processing the occurrence of multiple is always a great problem. Multiples have negative influences on velocity model building, seismic imaging, seismic attribute analysis (for example AVO), and structural interpretation, and so on.(1) SRME (surface-related multiple elimination) is a wave-equation-based multiple elimination method. Hereby, basic principles of SRME method were discussed in detail. The surface-related multiples are predicted in time domain and frequency domain respectively. Considering some problems in practice for SRME, the edge effect, wraparound problems, near-offset traces reconstruction by Radon transform, the application of source-receiver reciprocity and so on is studied in detail in order to achieve better result in this paper.(2) An in-depth theoretical analysis for MPI (multiple predicted through inversion) method, which was based on SRME, was completed in this paper. Synthetic data examples show that MPI is able to reduce the edge effect and partially explain the source signature, the detector patterns, and their directivity. The method can also predict internal multiples. Several iterations can improve the accuracy of prediction greatly.(3) The synthetic data and field data examples show that there are some time-difference between original multiple events and predictive multiple events, especially for subsurface with great lateral variation. A single-trace matching filter with time lag is proposed to correct the time difference, which better the subtraction of the predicted multiple.(4) The deliberate choosing of adaptive subtraction methods and the filter parameters are crucial to achieve better multiple attenuation result. Furthermore, iterative application of the filter can improve the result.Finally, the methods and strategy were used to process several synthetic data and field data. The multiple attenuation are satisfactory.