Improving Foz Do Amazonas Legacy NATS Data Using New Technologies in Seismic Imaging and Processing

In this work we demonstrate the benefits of reprocessing a legacy NATS dataset from 2014. We consider the water-bottom and the carbonate platform as the main sources of error in the imaging. The rugosity of the seafloor causes complex out-of-plane wave paths which require proper 3D algorithms, such as 3D designature and deghosting. It also causes strong multiple contamination which we address by employing a joint 3D surface related multiples elimination (SRME) and model-based water-layer demultiple (MWD) subtraction flow. The carbonate platform manifests as a strong lateral velocity variation that may cause distortion (pull-ups and push-downs) in the image. To build the model, an iterative workflow of Time-Lag Full-Waveform Inversion (TLFWI) combined with geological interpretation and ray-based tomography was applied. Finally, to overcome the azimuthal limitation of a NATS acquisition and to improve amplitude variation with offset (AVO), we applied a single-iteration Least-Squares Migration (LSM).