Hyperspectral and multispectral approaches for exploration and charactrisation of certain deposits of bauxite iron ore and limestone in South India

Abstract

This thesis is concerned with the development and application of certain recent remote sensing based approaches to locate and characterize iron ore, limestone and bauxite in the state of Tamilnadu, south India. The unique spectral characteristics of carbonate (limestone) and alumina (bauxite) and the existence of dedicated remote sensing sensors such as ASTER and Hyperion to map these minerals have encouraged the use of multispectral and hyperspectral remote sensing as a mineral exploration tool in this research. The methodology involves preparation of ASTER ratio maps such as ferrous mineral abundance, iron oxide abundance and ferrous silicate abundance to understand the distribution of iron rich components in Kanjamalai Hills. NDVI map is prepared followed by vegetation masking with the rescaled NDVI / DN values from 50 to 255 for vegetation. Image fusion using CARTOSAT-1 and ASTER images resulted in highlighting the additional/un-identified magnetite bands. There are 5 to 6 elliptical magnetite bands in the Kanjamalai Hill. The ferrous mineral abundance maps helped in highlighting the ferrous mineral present in the Kanjamalai hill and the magnetite ridges are clearly visible in the vegetation masked map. It is thus inferred that mineral mapping and grade assessment can be accomplished accurately using appropriate hyperspectral datasets and multispectral image analysis. Though improved spectral resolution certainly aids in identifying specific minerals based on their spectral response at narrow wavelengths, it is now inferred that a generalized methodology for mineral exploration cannot exist. Since the controls of ore-localisation differ from mineral to mineral, it is further inferred from this research that the approach to remote sensing based exploration will also vary according to the nature of ore-localisation. newline newline newline