Multidirectional Coherence Attribute for Discontinuity Characterization in Seismic Images

Abstract

Coherence attribute is an effective and widely used tool for identifying structural and stratigraphic anomalies in seismic data. By applying the coherence attribute on multiple bandwidth-, azimuth-, and offset-limited seismic data, multispectral, multiazimuth, and multioffset coherence attributes have also been proposed to enhance geologic discontinuities. Generally, geologic discontinuities, such as fault and paleochannel, usually have different dipping and meandering orientations and will be better identified along a perpendicular rather than a parallel direction in coherence computation. To address this issue, we propose a new multidirectional coherence attribute by combining an eigenstructure-based coherence algorithm with directional decomposition for discontinuity characterization. For 3-D seismic data, we compute eigenstructure-based coherence along multiple different directions to identify the seismic edges perpendicular to the chosen direction and set the minimum mean of multiple coherence images as the final multidirectional coherence attribute. Hence, the proposed multidirectional coherence attribute not only provides multiple coherence images highlighting partial edges in a certain direction but also offers a clearer coherence image highlighting whole edges in all directions than coherence without any decomposition. Furthermore, a colorful coherence image blending with several directional components intuitively contributes to interpret more geologic details. The coherence images of a 3-D physical modeling data and field data of carbonate reservoir prove the availability of the proposed method and demonstrate that multidirectional coherence attribute can serve as an improved tool to characterize the distribution of geologic discontinuity in seismic images.