The objective of the project Deep Electromagnetic Sounding for Mineral Exploration (DESMEX), funded by the Federal Ministry of Education and Research (BMBF), is the development of a semi-airborne controlled-source electromagnetic (CSEM) method. It combines powerful, grounded current transmitters with airborne magnetic-field receivers, providing higher signal-to-noise rations and area coverage than conventional land-based or airborne electromagnetic methods. With this methods, we are able to image the subsurface resistivity structure down to 1000 m depth.
The LIAG is primarily responsible for subproject IV of DESMEX – modeling and inversion. The main goal is to develop the open-source software custEM for the 3D finite-element (FE) modeling of semi-airborne CSEM data. For this measurement geometry, we need to be able to consider effects of topography, crooked transmitters and electric field receivers on the surface, and complex resistivity distributions (Figure 1). Furthermore, we are operating our powerful current-transmitter for the semi-airborne surveys, enabling source currents of approximately 20 A in 2 km long dipole transmitters. To be able to reference the results of our new method, we are also conducting large-scale electrical resistivity tomography (ERT) surveys and inverting the acquired data.
We established the 3D FE CSEM modeling toolbox custEM (Rochlitz et al., 2019; Rochlitz 2020), which meanwhile not only supports only semi-airborne CSEM simulations but also arbitrary airborne, land-based, marine and mixed geometries. The code is based on existing open source libraries. Most important, the finite-element kernel is based on the FEniCS project, which is steadily advancing. Hence, this library provides successively more useful features and better computational performance using state-of-the art algebra libraries and the most recent versions of direct solver MUMPS. The necessary unstructured tetrahedral mesh generation is conducted with TetGen, generating the required input file with our designed meshing tools based on the pyGIMLi library (Figures 2 & 3).
We implemented different total and secondary electric or magnetic field as well as gauged-potential approaches. The primary background-field solutions can be computed with high performance with the implemented routines in the simultaneously developed COMET toolbox. Interpolation and visualization tools simplify the post-processing workflow. The custEM toolbox is still under continuous development in the successor-project DESMEX II, funded as well by the BMBF. Meanwhile, many updates regarding robustness, performance and user-friendliness were implemented and the simulation capabilities were extended to support also time-domain and natural-source electromagnetic data.
As an example, we present one important application of the custEM toolbox in Figure 4. Here, we use the model in Figure 2 to show different measures to examine the differences in the field responses caused by the real topography in comparison to a flat-surface assumption. This result shows that in some areas, the relative misfits can easily reach more than 10 % and the magnitude field vectors can change their orientation up to 10° for this even comparatively smooth topography model. Therefore, it is necessary to consider the real geometry in the 3D semi-airborne data analysis and to develop forward-modeling tools like custEM which can handle such geometries. This is also a requirement for successfully inverting the survey data, which is our main goal in the successor project DESMEX II.
Further applications are presented by Rochlitz et al. (2019) and Rochlitz (2020). Moreover, the custEM source-code repository: https://gitlab.com/Rochlitz.R/custEM contains many examples how to use custEM for various electromagnetic modeling tasks.
The Schleiz area was the main survey region of the DESMEX project, where also other established electromagnetic methods such as airborne (HEM) and transient electromagnetics (LOTEM) were applied to reference our inversion results from the semi-airborne data. The main target was an antimonite deposit in this investigation area, which was exploited in the uppermost 100 m, but the trend to greater depths is unknown.
In the years 2015 & 2016, we conducted a 2D large-scale ERT survey with a total length of 7.5 km across the Berga anticline, near Schleiz, Germany (Figure 5, blue line, from Steuer et al., 2020). For our survey, we used a dipole-dipole setup with all in all 60 receiver locations and 125 m electrode spacing. The drive-in currents varied between 10 and 25 A. The final 2D inversion result of the processed DC data is shown in Figure 6 in a multi-method comparison with the other available data. For more details regarding the comparison, we refer to Steuer et al., 2020. The ERT on it’s own exhibits a significant resistivity distribution of the surrounding host rocks down to 500 m depth. The main conductive or resistive structures can be assigned to the main geological layers such as black-shales, greywacke or diabase.