The electromagnetic field affected by such inhomogenities propagates back to the surface (or seafloor) where it is recorded by the receivers. The recorded information is sufficient to create more or less detailed maps of the subsurface resistivity distribution. The recordings of electric and magnetic field contain a significant amount of natural noise, which along with equipment noise makes high resolution interpretation of EM data challenging. It is necessary to have a very detailed knowledge of how the energy propagates in the subsurface. In addition, advanced algorithms must be used to create the resistivity image.
Below you can see an illustration of modeling a range of scenarios to determine if targets can be identified using EM.

The key issue is integrated analysis of EM data and seismic/geological data. This includes numerous iterations during the EM processing/inversion to account for geological knowledge. The methodology described for marine environments, is a direct analogue to onshore environments. The integration of reservoir modelling and EM data carried out for hydrocarbons, is directly applicable to coal mining and production of coal bedded methane. The usefulness is well documented for petroleum applications, but there are potential applications within exploration of coal, ores and minerals. If successfully developed, EM-technology may become a cost-reducing technology with significant commercial potential. The current main objective is to develop the technology for distinguishing between sulphid-based ores and graphite in slate.

Source: Rocksource

