Development of a 2D adaptive finite element inversion algorithm for the marine Differential Electrical Dipole method

点击翻译按钮获取中文摘要

To increase the lateral resolution of a marine time-domain Controlled-Source Electromagnetic (TD-CSEM) application, the Institute of Geophysics and Meteorology (IGM), Cologne recently developed a novel seafloor-towed transmitter-receiver system called Differential Electric Dipole (DED). In the framework of a previous DFG project the novel DED method was developed and applied for the first time to study a sub-seafloor groundwater body near the coastline of Bat Yam, Israel. The application was effective in delineating the freshwater aquifer underneath the Mediterranean seafloor up to a distance of 3.6 - 3.7 km from the coastline.The acquired DED data set was interpreted using a 1D inversion approach, followed by a large-scale 2D modelling study that included over 400 000 forward calculations using the 3D finite difference forward algorithm for solving Maxwell's diffusion equations sldmem3t. Although the model parameter space was strongly simplified, the conducted calculations needed more than two months to complete. This interpretation approach is computationally expensive, biased, and needs to be replaced through the development of an efficient multi-dimensional (preferably 2D) TD-CSEM/DED inversion scheme.The proposed project aims to develop a fully functional finite element time-domain 2D inversion algorithm based on the existing frequency domain finite element forward algorithm of Li & Key (2007). The approach will be based on calculating the EM response for several frequencies in the frequency domain, followed by a transform into the time domain. The inversion scheme will include a global regularisation following Occam's razor to assure a smooth model inversion. The developments of the proposed project will justify future applications of the novel DED method in coastal regions to study freshwater aquifer systems. The 2D interpretation scheme is a necessary improvement for future applications, as DED exhibits an increased sensitivity towards multi-dimensional resistivity structures compared to the conventional HED source excitation. Furthermore, the bathymetry has a significant effect on DED data and needs to be sufficiently accounted for in the inversion scheme.