Capabilities of multispectral image in mapping and utilization of sensitive bands for characterization of waste rock dumps: A case study of a manganese mine in India

Abstract

Waste dump management is one of the major problems in an opencast manganese mine. The problem is further compounded as the stripping ratio is very high (1:20). The key approach of waste rock dump management is the reuse or reutilization of waste in form of aggregate or m-sand, which is also the basic tenet of circular economy. For waste reutilization, the role of multispectral satellite data analysis is considered to be the best approach in order to characterize the mineral content of the dump. The behavior of the spectral signature curves and its association with the surface reflectance (SR) values of multispectral data (Landsat 9 OLI) is used for determining the sensitive bands. Methods such as Object-Oriented Principal Components Analysis (PCA) and Independent Components Analysis (ICA) are used to study rock-mineral interactions. Formulas representing the optimal band ratio are also develop for precise mineral mapping. X-ray diffraction is used to identify the mineral constituents of the rocks. The integration of sensitive bands and incorporation of three methods (Band ratio, PCA and ICA) serve as the main objectives of the present study. The methods are applied to a opencast manganese ore mine in India to establish the sensitive bands for muscovite schists and gneissic rocks in the hanging wall and footwall face respectively. The spectral signature reflectance curves obtained from the multispectral data of the identified minerals are compared with the USGS spectral libraries to determine the sensitive bands (Bands 4, 5, 6, and 7) for rocks. False color composite images (FCC) in the RGB band combination (PC4-PC3-PC1) and (IC4-IC2-IC1) are utilized for precise mineral mapping of the study area. The methods are also validated (positive correlation R2 = 0.775) with the waste rock dumps of the same mine and also with the neighboring mine having similar geological origin.