Abstract We illustrate the use of curvature attributes for defining stratigraphic features of interest on horizons mapped in three-dimensional seismic data. Curvature is a two-dimensional property of a curve that quantifies how much the curve deviates from a straight line. Many different types of curvature may be defined for a surface, and these can be more useful than dip, azimuth or even 'conventional' (i.e. second derivative) curvature analyses for defining subtle structural or stratigraphic features of interest. In our workflow, we drape curvature over rendered three-dimensional surfaces and adjust lighting to highlight stratigraphic and structural features of interest. The examples we present are derived from clastic and carbonate settings of various ages, and include applications of curvature analyses to multibeam bathymetry and digital elevation model data.

Abstract The use of three-dimensional (3-D) seismic attributes to predict reservoir properties is becoming widespread in many areas. One of the most underutilized aspects of the methodology is that the property-prediction maps can help geoscientists understand depositional and postdepositional controls on reservoir development. We illustrate this point via a case study that examines partially dolomitized, restricted to open-marine carbonates of the Ordovician Red River Formation in the Williston Basin. We tied log and seismic data, mapped key reflection events in the 3-D seismic volume, calculated the porosity thickness (thickness × sonic porosity) for the porous zone, and then correlated those data with 21 attributes. We derived a relationship between two attributes (the spectral slope from peak to maximum frequency and the ratio of positive to negative samples) and porosity thickness that yielded a 0.88 correlation coefficient between predicted and actual values. This relationship was used to predict the porosity thickness throughout the 3-D seismic area. The resulting porosity distribution shows (1) good porosity development along the flanks of structures that are associated with visible faulting or steep dips at the underlying Winnipeg level, (2) thin (~17–28 ft [~5–8.5 m]) porous zones throughout much of the field, (3) a large, off-structure porosity zone in an area without well control, and (4) small, irregularly distributed porous zones (most likely the result of noise and/or error in the predictive relationship). In areas where faults and flexures are associated with enhanced porosity development, the slope of spectral frequency attribute may be responding to fractures, with more rapid attenuation of high frequencies occurring in these areas. These observations support a diagenetic model where faults and fractures acted locally as preferential pathways for dolomitizing fluids. Away from these zones, the porosity distribution shows some porosity thickness over the entire area that is consistent to drillstem test data that shows depleted pressures in wells drilled in the early 1990s on otherwise isolated structures.