Full text loading...
-
Multi-Dimensional Inversion-Driven Multiple Modelling for Sparse OBN Data
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 86th EAGE Annual Conference & Exhibition (Workshop Programme), Jun 2025, Volume 2025, p.1 - 5
Abstract
Summary
Demultiple of ocean-bottom-node data remains a challenging task in shallow water environments. Up-down deconvolution has become popular in such settings, producing an earth reflectivity estimate free of the source wavelet and surface-related multiples. The earth reflectivity may be taken forward for further processing or alternatively convolved with the input data to predict the corresponding surface-related multiples. The multiples may be directly subtracted or adaptively subtracted from the input data, a strategy which enables a thorough QC of the demultiple process. We propose an inversion-driven multiple prediction approach for ocean-bottom-node data based on source-side convolutions. The method may be applied with similar horizontally homogeneous geology (1.5D) assumptions to many up-down deconvolution implementations or alternatively using full 3D multi-dimensional summations, which overcome the flat-layer assumption. As normally the 3D approach is under-sampled, we propose a 2.5D implementation working on each receiver-line separately, with a translation of the prediction operator in the y-direction. The algorithm is supported by synthetic data analysis as well as a real dataset from the Norwegian North Sea.
© EAGE Publications BV