Publications

Three-dimensional (3D) seismic data are acquired in five-dimensions: source-x, source-y, receiver-x, receiver-y, and time. These five dimensions can also be represented by inline, crossline, offset, azimuth, and frequency. For narrow azimuth data, the dimension reduces to four, as azimuth is assumed to be nearly constant. Typical 3D seismic data always suffer from poor acquisition sampling along one or more spatial dimension. Five-dimensional (5D) interpolation techniques can interpolate simultaneously all space dimensions. Although interpolation is not a replacement of acquisition, nevertheless, 5D interpolation has proven to be quite successful in recent years. This is reflected by its use in increasingly challenging scenarios with more demanding requirements in various parts of the world. In this work, we evaluate the uplift using 3D versus 5D interpolation applied to the KE2 3D transition zone (95 km2). Acquired data are observed to be very sparse in near and far offset ranges and definitely needed higher dimension interpolation to produce an accurate interpolation of these offset ranges. This interpolation is applied pre-migration and the data are migrated using the Kirchhoff migration algorithm to compare the uplift on migrated images. To enhance the Kirchhoff migration output further, we add surface related multiple elimination (SRME) and source de-ghosting to the preprocessing flow (before 5D interpolation) followed by pre-stack time migration (PSTM). The SRME technique helps to attenuate surface multiples present in the data that contaminate migration results. Source de-ghosting minimizes the effect of source ghost and enhances the low frequency bandwidth. Pre-stack time migration helps to move the data to actual subsurface locations. In this case study, we show that the combination of all these recently developed techniques can produce a better seismic image than vintage processing.