Satellite Imagery for Monitoring and Mapping Soil Chromium Pollution in a Mine Waste Dump

Weathering and oxidation of sulphide minerals in mine wastes release toxic elements surrounding environments. As an alternative to traditional sampling chemical analysis methods, the capability proximal remote sensing techniques was investigated this study predict Chromium (Cr) concentration 120 soil samples collected from a dumpsite Sarcheshmeh copper mine, Iran. The samples’ mineralogy Cr were determined then subjected laboratory reflectance spectroscopy range Visible–Near Infrared–Shortwave Infrared (VNIR–SWIR: 350–2500 nm). raw spectra pre-processed using Savitzky-Golay First-Derivative (SG-FD) Second-Derivative (SG-SD) algorithms. important wavelengths Partial Least Squares Regression (PLSR) coefficients Genetic Algorithm (GA). Artificial Neural Networks (ANN), Stepwise Multiple Linear (SMLR) PLSR data mining methods applied selected spectral variables assess concentration. developed models bands Aster, Hyperion, Sentinel-2A, Landsat 8-OLI satellite images area. Afterwards, rasters obtained best prediction model segmented binary fitness function. According outputs spectroscopy, highest accuracy ANN SD with R2 = 0.91, RMSE 8.73 mg/kg RPD 2.76. SD-ANN also showed acceptable performance on mapping spatial distribution ordinary kriging technique. Using images, SD-SMLR provided values 0.61 0.53 for Hyperion respectively. This led higher visual similarity Sentinel-2A map. study’s findings indicated that applying by wavebands imagery could be considered promising technique rapid, cost-effective eco-friendly assessment highly heterogeneous areas.