Mapping the distribution of iron ore minerals and spatial correlation with environmental variables in hilltop mining areas

Abstract

The prime contribution of this assignment was to examine the hyperspectral remote sensing, based on iron ore minerals identification using spectral angle mapper (SAM) technique. Correlation analyses between field iron contents and environmental variables (soil, water, and vegetation) have been performed. Spectral feature fitting (SFF) and multi-range spectral feature fitting (MRSFF) methods were used for accuracy assessment in extracting iron ore minerals from Hyperion EO-1 data. Spectral inspections as a reference were used in SAM technique for image classification for iron ore minerals: Hematite (24.26%), Goethite (32.98%) and Desert (42.76). Iron ore minerals classification is justified by the United States Geological Survey (USGS) spectral library and field sample points. The regression analysis of USGS and Hyperion reflectance spectra has shown the moderate positive correlation. The regression analyses between iron ore contents and environmental parameters (soil, water, and vegetation) have shown the moderate negative correlation. The examination was significantly effectual in extracting iron ore minerals: Hematite (SFF RMSE ≤ 0.51 MRSFF RMSE ≤ 0.48), Goethite (SFF RMSE ≤ 0.047 MRSFF RMSE ≤ 0.438) and Desert (SFF RMSE ≤ 0.63 and MRSFF RMSE ≤ 0.50); and the MRSFF RMSE histograms indicate the above result likened to a conventional SFF RMSE. MRSFF RMS error result is best because multiple absorption features typically characterize spectral signatures. This analysis demonstrates the potential applicability of the methodology for iron minerals identification framework and iron minerals impact on environmental parameters.