Shallow-water ocean-bottom node demultiple using joint wavefield extrapolation multiple modeling and surface-related multiple elimination with subtraction in the curvelet domain

Ocean-bottom node (OBN) surveys are increasingly used for seismic exploration, as OBN geometries provide long-offset and full-azimuth illumination, relative to traditional towed streamer surveys. This in turn facilitates more accurate velocity model building and imaging. Surface-related multiple elimination (SRME) is a well-established method for predicting surface-related multiples in OBN data; however, it has limitations due to the absence of near-offset data and the subsequent lack of reliable water-bottom reflections in shallow-water environments. In addition, amplitude overprediction of higher order of multiples is a known challenge for SRME is shallow water environment due to first-order approximations when convolving the data with itself. Wavefield extrapolation multiple modeling (WEMM), a model-based approach, addresses these SRME limitations by deterministically predicting all surface multiples over all offsets via one-way wavefield extrapolation. As OBN is increasingly the choice of acquisition technology, effective and efficient demultiple processing solutions are becoming increasingly critical. In this study we present a joint WEMM and SRME approach applied to a shallow-water OBN survey. High-frequency WEMM is employed to efficiently handle shallow-water bottom surface multiples, while remaining surface-related multiples are addressed by SRME. The methodology was optimized for both frequency and resolution, resulting in a high-resolution, multiple-free OBN dataset.