Multidimensional multiple attenuation in complex geology: illustration on the Sigsbee2B dataset

Antoine Guitton

antoine@sep.stanford.edu

ABSTRACT

A pattern-based multiple attenuation technique is tested on the Sigsbee2B dataset. The results of multiple removal are analyzed in the image space after migration to better understand the impact of this process on the primaries. When an accurate model of the primaries (signal) and the multiples (noise) exist for the estimation of the noise and signal annihilation filters (i.e., non-stationary multidimensional prediction-error filters), this method removes the multiples extremely well while preserving the primaries. When the model of the multiples is provided by the Spitz approximation, which consists in convolving the data with the noise annihilation filters to get a signal model, good results are obtained if 3D filters are utilized.