Waveform Inversion-Based Internal Multiple Suppression by Constructing Virtual Events

Abstract Internal multiples severely reduce seismic imaging quality and interpretation accuracy. With the continuous advancement of seismic imaging toward higher resolution, the accurate suppression of internal multiples has become increasingly critical. The concept of virtual events provides a fully data-driven approach for suppressing internal multiples. However, when energy interference occurs between internal multiples and primaries, the process of adaptive subtraction using matching filters may lead to damage to primaries. Building upon the conventional virtual events theory, an adaptive iterative method for the internal multiple suppression had been developed, which reduces the dependency on the matching algorithm to some extent. Nonetheless, it remains unable to fundamentally prevent the loss of primary energy. To address this issue, we propose a waveform inversion-based internal multiple suppression method by constructing virtual events. Based on the theoretical foundation of the traditional virtual-events theory, we formulate an inversion objective function to estimate internal multiples via waveform inversion, thereby avoiding the requirement for the adaptive subtraction process. This approach achieves effective suppression of internal multiples while better preserving the primary energy. Comparating results with traditional methods, we demonstrate that the proposed method exhibits superior performance in both accuracy of multiple suppression and preservation of primary amplitude, providing a more reliable data foundation for subsequent seismic processing and geological interpretation.