Attenuation of Multiples in 2D Short Streamer Length Multi-channel Seismic Data, Gulf of Callifornia

Abstract

Marine seismic data are typically imprinted with multiples energy, which is undesirable in seismic imaging. Due to the coherent nature of multiples, even if the data are long-offset, they are difficult to remove. Here, we present a comparison of four methods for attenuating long-period multiples energy in short-offset 2D streamer data from Gulf of California. Multiple attenuation methods include deconvolution, Linear Radon (Tau-P), Parabolic Radon Transform, and 2D Surface Related Multiple Elimination (SRME). Multiple attenuation entailed selection of most apt method or combination of methods for optimum results, without compromising quality of the primaries in the data. Results suggest that any kind of deconvolution based approach is the least effective. The Linear Radon filtering is effective in selective parts of the model. Both Parabolic Radon transform and SRME appear to be the most effective methods, but not free of inherent pitfalls. The sub-seafloor velocity model appears to be playing a key role in the Parabolic Radon Filtering. We conclude that while there is no general rule in processing multiples attenuation, tailoring methods based on acquisition and subsurface geology often produces the most acceptable result. In our case, we interpret the stack that was produced after application of SRME method. The resulting stack shows new magmatic bodies that were suppressed by multiples energy.