Full text loading...
Evolving from acoustic FWI (aFWI) to elastic anisotropic FWI (eFWI) of long-offset /wide-azimuth 4C OBN data aims at narrowing the gap between imaging and reservoir characterization. Two challenges are the computational burden of accurate elastic wave modeling and multiparameter reconstruction from 4C data. While aFWI is generally performed with finite-difference schemes, ultra-low shear wavespeeds (Vs) near the sea bottom may require to perform elastic modeling on unstructured meshes to achieve the best trade-off between accuracy and computational cost. In this context, we assess the spectral element method (SEM) on unstructured hexahedral meshes as forward engine of time-domain eFWI with the Gorgon 4C-OBN dataset, Australia. We apply eFWI on a low-frequency narrow-azimuth subset of the pressure data to update compressional wavespeed (Vp), Vs and density. Multiparameter inversion leads to better convergence than monoparameter inversion for Vp. The good data fit achieved after eFWI, the structural consistency between the long-wavelength elastic FWI model and a legacy 25 Hz acoustic FWI model, and the consistent Vp/Vs ratio in the reservoir with empirical relationships for sand and shale lithologies validate our SEM-based elastic FWI. Ongoing work aims to evolve toward the inversion of the four components for the reconstruction of the full elasticity tensor.