Improved Up-Down Deconvolution in Ultra-Shallow Waters Offshore Abu Dhabi

Abstract

Abstract The ultra-shallow water-depth and hard water bottom of offshore Abu Dhabi are responsible for generating a complex multiple wavefield. Removing this energy is critical for accurate imaging of the subsurface. Almheiri et al. (2022), demonstrated the effectiveness of up/down deconvolution to attenuate multiple energy on Ocean bottom sensor (OBS) data from offshore Abu Dhabi. In this paper we detail recent advances to this workflow, leading to enhanced imaging results from offshore Abu Dhabi. Amundsen (2001) showed that deconvolving the up-going wavefield by the down-going wavefield produces an estimation of the Earth's reflectivity series. While this technique gives a step-change in demultiple results from OBS data offshore Abu Dhabi, it suffers from high noise levels – particularly in the higher frequencies. Advances in linear noise removal enable the attenuation of mud-roll from the input data without harming the water-wave – an essential ingredient to up/down deconvolution. Removal of the mud-roll consequently enables better attenuation of shear-wave noise from the vertical geophone component. The advanced workflow leads to improved wavefield separation and superior data for up/down deconvolution.The advanced pre-processing flow was applied to an OBS dataset from Offshore Abu Dhabi and used as input to up/down wavefield deconvolution. The up/down deconvolution results were benchmarked against a previous result from the same OBS dataset, without the application of the improved pre-conditioning flow. Reverse time migration (RTM) imaging using both datasets showed a clear improvement in signal-to-noise ratio on the new results, with improved reflector focusing, especially at the reservoir target level. Frequency band splitting showed better phase alignment across the bandwidth in the new data compared to the legacy results. These results confirm that advanced de-noising techniques, allowing for the removal of high-amplitude mud-roll and shear-wave noise, without harming the water-wave, refraction energy or primary reflections, leads to improved wavefield separation and consequently improved up/down deconvolution results. The improved results exhibit reduced noise content, better event focusing and improved phase alignment across the frequency spectrum.