Demultiple strategy combining Radon filtering and Radon domain adaptive multiple subtraction

摘要

The traditional cascaded demultiple strategy exploits the time-offset domain adaptive multiple subtraction and parabolic Radon transform method sequentially to suppress multiples. The time-offset domain adaptive multiple subtraction may cause residual multiples, especially for multiples overlapping with primaries. If these residual multiples have residual moveout close to that of primaries, they cannot be removed effectively by the traditional cascaded demultiple strategy. In this paper we propose the demultiple strategy combining Radon filtering and Radon domain adaptive multiple subtraction. We divide the Radon image of the original data into three areas. Wemute the Radon area where primaries map into, and keep the Radon areawhere ineffectively predicted multiplesmap into. For the third areaweuse the matching filter to estimatemultiples. The three Radon areas are then blended into one Radon image, which is transformed into the time-offset domain to model multiples. Finally the modeled multiples are subtracted fromthe original data to obtain estimated primaries. The proposed demultiple strategy can better separate primaries and multiples with close residual moveout in the Radon domain than the traditional parabolic Radon transform method and cascaded demultiple strategy. Tests on synthetic and real data sets demonstrate the effectiveness of the proposed demultiple strategy.