Enhancing coherency analysis for fault detection and mapping using 3D diffraction imaging

Automatic detection of geological discontinuities such as small throw faults, and pinch-outs is an important problem in the interpretation of 3D seismic data. This is commonly done using coherency analysis. However coherency may be affected by noise, which may create false anomalies. We propose a new interpretation workflow for the detection and mapping of faults, which enhances the coherency-type analysis with identification and detection of diffractions produced by the discontinuities. The algorithm utilizes migrated and unmigrated stacked seismic volumes and the cube of stacking (NMO) velocities. Tests on a simple 2.5 D model show that the method is capable in detecting and mapping of faults below seismic resolution.