The climate crisis and the need for a rapid transition to renewable energy systems have contributed to a pressing demand for critical raw materials. The wastes of historical and abandoned mines have been targeted for exploration of such minerals, but many of these sites are also affected by mine drainage, which pollutes nearby soils and waterways. Hundreds of thousands of historical abandoned mines are distributed across Europe, and with them large volumes of mine residues. For several of these, the chemical composition is still unknown. UAV, aerial and satellite imaging spectroscopy are effective means for monitoring acid mine drainage and the associated secondary mineral assemblage. However, despite still being a significant source of pollution due to the mobilization of large amounts of heavy metals, few studies have investigated slightly acidic to circum-neutral mine drainages using a remote sensing approach.

In this context, the present study aims at identifying sources of mine drainage in the historical carbonate-buffered Pb-Zn mine wastes of Montevecchio and Ingurtosu, in South-West Sardinia, by using multispectral and hyperspectral remote sensing. Soil and tailings samples were collected in two field throughout the year 2024 and were measured spectrally with an ASD FieldSpec-3 spectrometer in the lab. The major minerals within the samples were identified with an iterative polynomial fitting technique. Maps of the dominant iron-bearing minerals, present as secondary minerals, were extracted by applying a polynomial fitting technique on three Sentinel-2 images acquired in different seasons of the year 2024 and concurrent with the field campaigns and a hyperspectral EnMAP image. The results obtained from the satellite images were validated with the field sample spectra.

EnMAP and sample spectra reveal a spatial zonation of Fe(III)-bearing hydroxides and sulfates, highlighting potential hotspots of acid mine drainage generation in mine residues. Moreover, subtle compositional changes of Zn and Fe-bearing efflorescent minerals with pH variations could be observed along a river affected by slightly acidic to circum-neutral drainage. With Sentinel-2 data it is possible to distinguish different major ferric iron minerals such as goethite, hematite and jarosite, though the quality of the results is highly sensitive to atmospheric correction and depends on prior knowledge of the area’s geology. The temporal changes of the maps illustrate the variation of relative abundance of such minerals along with environmental conditions such as precipitation and humidity. This work shows the potential of a multiscale and multiplatform approach can be a powerful tool for an in-depth characterization of mining site from a geochemical and mineralogical point of view, setting the baseline for further applications in mining and post-mining contexts.