A New Method for Diffraction Separation based on Trace Randomized Realignment and Denoising in Migrated Dip-Angle Domain

Abstract

Summary Diffractions occur when seismic waves strike on lithologic or structural discontinuities in subsurface and they are therefore useful for the prediction of geological abruptions. Due to the weakness of diffractions compared with reflection energy, diffraction separation is required. Migrated gathers in dip-angle domain reveal a clear difference in shape between reflection and diffraction events. According to the distinction, we put forward a novel diffraction separation method called trace randomized realignment and denoising, the core concept of which is to realign the traces in each raw migrated dip-angle gather randomly, eliminate the reflections in the gathers as noise and preserve the diffractions. Synthetic and real data examples confirm the validity and accuracy of the proposed method from the perspective of dip-angle gather as well as diffraction image.