Marine CSEM 2D inversion guided by the eigenvectors of seismic image

Marine controlled-source electromagnetic (CSEM) inversion can produce useful resistivity models for offshore hydrocarbon surveys. However, traditional inversion techniques often face a trade-off between stability and resolution in resistivity imaging. To improve the imaging quality, we develop a novel image-guided regularization technique that works on an unstructured mesh. This approach adjusts the model penalty intensity following the eigenvectors derived from available seismic images, considering both the relative positions and features between adjacent cells. This modification guides the reconstructed resistivities to align with the seismic structures. We evaluate its effectiveness using the flattest model and the minimum gradient support (MGS) constraints within the framework of Occam’s inversion. Both synthetic and field data examples from the Troll West Oil Province (TWOP) demonstrate that the image-guided inversion effectively unveils subsurface resistivity structures and delineates the true shape and lower boundaries of resistive targets. The robustness and high-resolution imaging capability of the proposed method underscore its potential as a promising tool for offshore reservoir exploration.