Structurally Constrained Anisotropic Multi-Wave-Inversion Utilizing Machine Learning and Big Data on a Middle East OBC Project

Summary Challenged by the presence of strong anisotropy and velocity reversals in the near surface, we apply structurally constrained anisotropic multi-wave inversion (MWI), over 1200 km2 of OBC data offshore Abu Dhabi. MWI aims to simultaneously invert the P-wave first breaks (FB), the ground roll dispersion curves (DC) and the vertical two-way times (VTWT) of the main interfaces. In this study, MWI is extended to invert for the Thomsen’s anisotropic parameter  that is constrained by the FB and the VTWT. The high resolution Vp and  obtained from MWI greatly contribute to the velocity model building. Additionally, geomechanical parameters like the uniaxial compressive strength (UCS) are also derived from the combination of Vp and Vs. We highlight the importance of proper preconditioning of the DC and FB inputs using data mining tools when applying the method to a large production dataset. Machine learning (ML) is used to better account for the vertical and horizontal geological variability of the survey in picking the phase velocity, while data mining tools allow interactive QC and editing of the FB picks on large data sets.