The value of streamer data in complementing a sparse OBN survey for elastic FWI imaging

The fidelity of full waveform inversion (FWI) results depends on several factors, such as appropriate physics, adequate sampling, sufficient low frequencies, and a high-quality starting model. Sparse long-offset low-frequency (LOLF) ocean bottom node (OBN) surveys have proven effective in building the low wavenumbers into velocity models, thanks to their ability to provide low frequencies with adequate sampling, superior long offset coverage and signal-to-noise ratio (S/N). However, sparse LOLF OBN surveys suffer from imaging limitations at high frequencies due to the survey design characteristics. In such scenarios, dense datasets (streamer or conventional OBN) can be used to overcome these challenges. In this paper, a case study is presented where a sparse LOLF OBN survey using the Tuned Pulse Source (TPS) is complemented with legacy streamer data for visco-elastic FWI (eFWI) imaging in the Gulf of America (GoA). Comparisons between OBN-only, streamer-only, and joint OBN and streamer eFWI demonstrate the utility of streamer data beyond 7 Hz for adequate eFWI imaging. We conclude that a staged FWI approach with the sparse OBN providing the low wavenumbers, followed by a joint inversion of the OBN and streamer data produced the best result up to 10 Hz.