5D Compressive Sensing Conditioning with Dual Coil Shooting Data
W. Sanger, A. Zarkhidze, M. Vassallo and P. Vascik
Event name: 81st EAGE Conference and Exhibition 2019
Session: AI-Driven Interpolation
Publication date: 03 June 2019
Info: Extended abstract, PDF ( 2.96Mb )
Price: € 20
We present a new workflow applied to a dual-vessel marine seismic streamer acquisition using a circular-shooting geometry to extract and improve valuable information for the imaging process. In this work, we focus on the results for subsalt imaging. This approach allows us to simultaneously interpolate, regularize, and improve the signal-to-noise ratio in a 5D transform domain while taking benefit from the compressive sensing nature of the acquisition geometry. Compressive sensing (CS) enables reconstructing the signal from seismic data beyond the traditional Shannon-Nyquist criterion. CS requires two elements: (1) incoherent sampling (generally random), and (2) representation of the signal in a transform domain where it is sparse. The transform can include the Fourier transform, Radon transform, curvelet transform, or even the migration operator. A dual-vessel circular-shooting geometry with random sampling in four dimensions has seen broad use in the complex salt environment of the Gulf of Mexico. We apply compressive sensing conditioning (CSC) to circular-shooting, full-azimuth data to explicitly promote the sparse low-to-mid-frequency coherent signal, which is valuable for imaging subsalt targets. We show benefits to model building and far-offset imaging with the CSC process.