Internal multiple elimination based on the focusing concept — Comparison of common-focus-point layer-related method and Marchenko redatuming-based method

ABSTRACT A focusing theory plays a crucial role for internal multiple elimination (IME). The “focusing” concept involves simulating an experiment with sources and receivers placed at depth in a medium that extends infinitely as a homogeneous medium above the new acquisition surface and matches the real medium below the acquisition surface. An early implementation of the focusing concept is the common-focus-point (CFP) technique, which accounts only for primaries and provides insights into raypaths, events, and subevents via downward continuations, whereas the focusing in the recently developed Marchenko theory can account for internal multiples. Both theories share a focusing concept, but distinct focusing processes lead to differences in redatuming-based IME. A comprehensive comparison of the CFP layer-related method and a Marchenko method, namely the redatuming-based method, for IME is conducted. Investigating the theoretical descriptions of both methods indicates that intermediate steps involved in both methods exhibit similar behavior once the layer-stripping approach is adopted. These steps include the computation of CFP gathers versus upgoing Green’s functions, convolution of CFP gathers with gridpoint gathers versus downgoing Green’s functions, and prediction and elimination of the corresponding multiples. The focusing process in the CFP method can be considered as a one-way approximation of the Marchenko method. Unlike conventional layer-stripping approaches, the Marchenko redatuming method can focus on primaries and multiples, enabling a more flexible redatuming process. Furthermore, conclusions are drawn regarding influencing factors in both methods. Boundary selection is critical for the accuracy of both methods, and both exhibit a certain robustness to the macro velocity model. The findings observed in the 1.5D case are validated in a more complex 2D inhomogeneous setting, demonstrating their broader applicability. Revealing the connections between these two methods provides valuable insights into the focusing theory and the IME methods.