Methods for isolating coherent noise in the Radon domain

Shauna Kaye Oppert, R. James Brown

Discrete Radon transforms are often employed for the discrimination of reflections having parabolic or hyperbolic moveout, with no variation in amplitude with offset. The present work has tested the effects of data weighting and variants of the Radon transform for primary, multiple, and mode-converted reflections on a flat-lying model. Data weighting prior to the application of a Radon transform allowed for accurate identification of mode-converted reflections from other events. The pre- and postcritical energy from the mode-converted reflection was successfully separated in the Radon domain through application of a multiple-hyperbolic and a fourth-order nonhyperbolic Radon transform on separate near- and far-offset panels. Focusing of events in the Radon domain was maximized when we used a t2-stretched parabolic transform or a fourth-order nonhyperbolic transform tuned for mode-converted reflections. Additional work is necessary to improve the robustness of the algorithms.