Common reflection point stacking for ocean-bottom node data

Abstract

Due to the asymmetrical ray paths caused by the large elevation differences between source and receiver locations in ocean-bottom node (OBN) survey, conventional common midpoint (CMP) stacking is unsuitable for processing OBN data. To mitigate this problem, we propose a common reflection point (CRP) stacking method, which enables us to efficiently produce accurate CRP-stacked image and stacking velocity model for subsequent processing workflow. For each input trace, we estimate the samples representing the reflection points in the output time section using stacking velocity, and then use a modified double-square-root (DSR) formula to map the amplitudes of input trace to these samples, which are progressively accumulated to form moveout-corrected CRP gathers. We can save computational costs by restricting the samples that need to be processed based on OBN reflection geometry, and use conventional moveout analysis of the CRP gathers to iteratively update the stacking velocity model. With tests on synthetic and field-data examples, we demonstrate the advantages of the proposed CRP stacking over conventional CMP stacking.