A method for adaptive subtraction in conjunction with internal multiple prediction by inverse scattering series is investigated. An L 1 , nonstationary adaptive subtraction is found to minimize mistaken matching to primaries while still allowing for a large degree of multiple removal. This adaptive subtraction is tested on both synthetic and physical modeling data. The method proves to be effective, and avoids unwanted matching to primaries.
View full article as PDF (1.02 Mb)