Full Waveform Inversion Using Broadband Variable-depth Streamer Data in Deep Water West Africa

Full waveform inversion (FWI) produces high-resolution velocity models through an iterative inversion scheme that minimizes the difference between real and modeled data. Due to the limitations of current FWI algorithms, clean, low-frequency data are desirable to minimize the potential for cycle skipping and crucial for a successful FWI update. We apply FWI to a high-quality, broadband, variable-depth streamer data set acquired in the deep water Kwanza Basin of offshore Angola. We jointly update V0 and ε at low frequencies, and then follow with a V0-only, higher-frequency inversion. The jointly updated V0 and ε models show improved resolution and improve image and gather quality. The high-frequency update further refines the velocity model by detecting smaller scale velocity variations that are unresolvable with a low-frequency inversion. We describe our workflow and present the results of the joint V0 and ε inversion and the higher-frequency, V0-only inversion.