Separation of Multi-Source Blended Seismic Acquisition Data by Iterative Denoising

Han Li-Guo,Tan Chen-Qing,Lu Qing-Tian,Zhang Ya-Hong,Gong Xiang-Bo

DOI: https://doi.org/10.6038/cjg20130726

2013-01-01

Chinese Journal of Geophysics

Abstract:In multi-source blended seismic acquisition, different sources are shot in an overlapping fashion with certain time delays and blended records are acquired, so that both acquisition efficiency and image quality can be significantly improved. Deblending is the procedure of recovering data as if they were acquired in the conventional survey. A simple least-squares procedure is only able to get pseudodeblend results, where the blending noises cannot be removed. In high blending factor data, the blending noises are usually several times higher than the useful signals, which multiply the difficulty of source separation. Fortunately in pseudodeblend records, these noises are only coherent in the common source domain, but incoherent in other domains. For this character, multilevel median filter and Curvelet threshold iteration denoising are jointly used in this paper, and a new source separation method based on the iterative denoising in different domains is introduced. Depending on the different character of blending noises in different domains, corresponding denoising method is utilized, and an iteration method is designed for the optimization. While dealing with numerical blended real dataset, ideal results could be produced after only a few iterations, which verify that our method can largely improve the separation quality and calculation efficiency.