Fast-track, Data-driven Interbed Multiple Removal: Application to the Gulf of Suez

By

 Roald Gunnar van Borselen¹, Grog Fookes¹, Ian M. Threadgold², Norman C. Allegar²

(1) PGS Geophysical, Walton-on-Thames, United Kingdom (2) BP Egypt, Maadi, Cairo, Egypt

 Removal of multiples from seismic reflection data is an essential pre-processing step before seismic imaging in many marine environments. Surface-Related Multiple Elimination (SRME) has proven to be a valuable tool to remove free-surface multiples, multiples that have propagated down- and up-the-water layer more than once. This paper discusses the application of SRME to remove multiples generated by internal surfaces.

There are two generic strategies to tackle interbed multiples: model-driven methods, that make use of statistical assumptions and/or a priori information about the subsurface (local 1D assumption, detailed velocity- and/or reflector information), and data-driven methods, that use the measured data itself to predict and subtract interbed multiples.

Although conventional, model-driven approaches have been applied successfully, the reliability of the inherent assumptions and the user-provided priori information, as well as the level of user-interaction required makes these methods less suitable for large 3D production processing.

The approach taken here is an extension of the SRME method and requires only the identification (i.e. picking) of the multiple generator. Then, through a double convolution of muted common receiver- and common shot gathers, 2D internal multiples can be predicted. Through least-squares subtraction, these predicted multiples are subtracted from the input data.

An optimised processing strategy leads to efficient removal of either interbed multiples that are generated by a (chosen) internal reflector, or interbed multiples that have crossed a (chosen) pseudo boundary during wavefield propagation.

Application to a data set from the Gulf of Suez leads to encouraging results.