Paper Summary

A novel, de-aliased method called Stepwise Multiple Elimination using Linear Transforms (SMELT) is introduced as an alternative to 2-D Surface Related Multiple Elimination (SRME) to remove problematic multiples on the near-to-mid offsets of common midpoint (CMP) gathers in the south-western Barents Sea. A locally strong Base Quaternary reflector contaminates the SRME multiple model with severe amplitude errors that propagate through the underlying and prospective Triassic-Permian sequence. Adaptive subtraction of this model in any domain leads to the deterioration of primary energy in the target interval. No such deterioration is seen after the application of SMELT, which when used in conjunction with other demultiple processes effective on the mid-to-far offsets, allow the full offset range to be stacked.