We extend applications of a recently developed method for internal multiples prediction (van der Neut and Wapenaar, 2016) to a streamer dataset from a shallow-water survey. The internal multiples being predicted are all the internal multiples arriving later than a reflection from a fictitious reflector that separates the subsurface into overburden and target areas. The predicted internal multiples include all the internal multiples that interfere potentially with primaries from the target area. Similarly to other data-driven prediction methods, this new method combines sub-events in the data by convolutions and crosscorrelations and applies specific muting operations to the input data and to intermediate results in the prediction process. Using synthetic and field data examples we illustrate the effectiveness of this method to predict all internal multiples for a deep target area without requiring any subsurface information about the generators of these multiples. We also point out challenges for this new method, for instance for predictions at large offsets or for shallow targets. Finally, we discuss potential strategies for improving the predictions of internal multiples by this approach for the case of shallow-water surveys.
Presentation Date: Monday, October 12, 2020
Session Start Time: 1:50 PM
Presentation Time: 1:50 PM
Location: Poster Station 13
Presentation Type: Poster
