Noise is often the dominant problem with land seismic data and very few new processing techniques have been developed in the last 5 years or so to address this. Acquisition advancements have been made with deployments of more receivers and compressed sensing technology, but data quality is still limited by shallow subsurface noise, particularly scattering. We show here how recent new algorithms that have shown huge improvements in data quality in other parts of the world have been used to build confidence in vintage data from the Beetaloo Sub-basin. Some 2,000km of multi-vintage 2D data (1989-2012) covering much of the Beetaloo Sub-basin was reprocessed through a modern workflow aimed at reducing the impact of the noise and creating a unified dataset fit for delineating structural geometries and planning future drilling and seismic work in the area. Key elements of the new workflow discussed here include Wave Equation Refraction Statics; Full Waveform Corrections; Denoise in various domains - Radon, FK, Cadzow, etc. The combination of these technologies has resulted in a huge uplift in data quality and a unified dataset, enabling planning for future activities to be made with greater confidence than was possible using previous, disparate datasets and assist in future development of a highly prospective basin.

Open-Access Online Publication: May 29, 2023