Publications

Multiple elimination is one of the fundamental problems in seismic data processing and is still an active area of research. Numerous methods have been proposed for the elimination of multiples, each depending on a specific criterion for discriminating between primaries and multiples. One basic approach utilizes moveout properties that differentiate primaries from multiples, for example, f-k and Radon demultiple filters. Other methods predict and subtract multiples from seismic data. Examples of such methods are predictive deconvolution and the more recently developed wave-equation based methods.A comparative study of Tau-p predictive deconvolution, Radon demultiple, and ARCOs proprietary Wave Equation Multiple Suppression (WEMS) method was conducted. We present the results of applying these three multiple suppression techniques to a data set with a moderately hard water bottom and lateral structural variation.The effectiveness of both Radon and Tau-p deconvolution methods depends on the amount of structure, the subsurface velocity profile, the hardness of the seafloor and the water depth. While WEMS does not depend on any subsurface properties, it is sensitive to factors that can affect the accuracy of the near offsets, e.g., shallow water depths. Since each method has different strengths and weaknesses, a combination of these three methods can be used to optimally attenuate multiples in any kind of data.Since the example data set contained significant structure, WEMS provided the most incremental multiple suppression.