Computing Volumetric-Curvature Attributes Using Predictive Painting

Delineation of faults, fractures, and other discontinuities, either structural or stratigraphic, is crucial to seismic interpretation. Successful use of curvature attributes in the prediction of faults, fractures, and other stratigraphic features has proven curvature to be a powerful tool. The advent of volumetric estimation of curvature has exploited the curvature attribute to its fullest potential by overcoming the limitations of horizonbased attributes. We introduce a novel approach to computing volumetric curvature. The key idea is to transfer seismic image into a coordinate frame in which geometry follows the natural shape of each reflector, therefore assigning a horizon to each point in the seismic data volume is tractable. We employ the predictive-painting algorithm to estimate horizon shapes and to extract them automatically in the whole volume of seismic data. This computation can be used to transfer seismic data into two different coordinate systems that follow from predictive painting. The adopted curvature equations are defined in these two coordinate systems to compute curvature at each point in the volume. The proposed approach also enables multispectral curvature computation and capturing of curvatures at different scales. A field-data example shows that the proposed method can successfully estimate multispectral curvature of reflectors volumetrically.