A field data example of Marchenko multiple elimination

Abstract

Multiple reflections are often considered as noise in seismic data. Dealing with them has attracted much attention from industry and academia. A variety of schemes have been developed for removal of free-surface multiple reflections and success has been achieved in numerical and field datasets. However, less attention has been directed toward internal multiple reflection removal. The Marchenko multiple elimination scheme is capable to eliminate internal multiple reflections without model information or adaptive subtraction. This scheme is derived from coupled Marchenko-type equations by projecting the focusing functions from the focusing level back to the acquisition surface. Primary reflections are filtered from acquired seismic data. We apply the Marchenko multiple elimination scheme to a deep water 2D field dataset from the Norwegian North Sea to test the performance. We show that the Marchenko multiple elimination scheme is an appropriate method for removing internal multiple reflections when high-quality pre-processing is performed.

Wave-field representations with Green's functions, propagator matrices, and Marchenko-type focusing functions

“…These representations, transformed to the time domain, form the basis for the development of Marchenko schemes, aiming at resolving the focusing functions F 1 and F 2 from the reflection responses R 12 and R 22 at the acquisition boundary ∂D 0 . Such schemes have been successfully developed for precritical acoustic data (Ravasi et al 2016;Staring et al 2018;Jia et al 2018;Brackenhoff et al 2019;Pereira et al 2019;Zhang & Slob 2020;Staring & Wapenaar 2020).…”

Pre-processing methodology to apply the Marchenko method in conventional marine datasets

“…The Marchenko method is a powerful geophysical tool that allows us to build high-resolution images in depth and remove internal multiples from a seismic dataset (Wapenaar et al, 2014;Zhang and Slob, 2020). The quality of the results depends on the input field data preprocessing.…”