Discriminating Uranium and Copper mills using satellite imagery

Identifying uranium mills from high resolution commercial satellite images has assumed significance in recent years because of non-proliferation concerns. Studies have shown that it is difficult to identify Uranium mills through remote sensing methods that use only spectral signatures. In this communication we suggest an approach that relies only on spatial signatures of the equipment used in the extraction process as an alternative. Since the extraction of Uranium and Copper have many similar features especially where Copper is extracted from low grade ore or from copper tailings, there could be ambiguity in identifying a Uranium mill from high resolution commercial satellite images. In an earlier work carried out by the authors and summarized in this paper as well we had proposed a separation between copper and uranium mills based on the spatial signatures of equipment that is unique to the copper milling process. In this paper we suggest some improvements to the methodology outlined by us in our earlier work. In addition to the other features used to separate Uranium and Copper mills we bring in the dimensions of common equipment used in both processes as an additional dimension to improve the robustness of our classification. This technique is applicable only where the extraction is done in a mill and not where Uranium is extracted by in situ leaching methods.