SPECIAL SECTION: HYPERSPECTRAL REMOTE SENSING CURRENT SCIENCE, VOL. 108, NO. 5, 10 MARCH 2015 879 *For correspondence. (e-mail: ramakrish@iitb.ac.in) Hyperspectral remote sensing and geological applications D. Ramakrishnan* and Rishikesh Bharti Department of Earth Sciences, Indian Institute of Technology Bombay, Powai, Mumbai 400 076, India This article reviews the potential of Hyperspectral Remote Sensing (HRS) technique in various geological applications ranging from lithological mapping to exploration of economic minerals of lesser crustal abundance. This work updates understanding on the subject starting from spectroscopy of minerals to its application in exploring mineral deposits and hydro- carbon reservoirs through different procedures such as atmospheric correction, noise reduction, retrieval of pure spectral endmembers and unmixing. Besides linear unmixing, nonlinear unmixing and parameters attributed to nonlinear behaviour of reflected light are also addressed. A few case studies are included to demonstrate the efficacy of this technique in different geological explorations. Finally, recent developments in this field like ultra spectral imaging from unmanned aerial vehicles and its consequences are pointed out. Keywords: Geological applications, hyperspectral remote sensing, spectroscopy of minerals and rocks, spectral unmixing. Introduction MINERAL exploration and geological mapping through conventional geological techniques are tedious, expensive and time-consuming. Mapping and targeting an economic deposit through traditional techniques involves extensive fieldwork, structural mapping, study of landforms, petrography, mineralogy and geochemical analyses 1,2 . These techniques need a strong laboratory database to discern slight variation in composition of ore grades. With the advent of multispectral sensors (e.g. ASTER, Landsat) having bands in the Shortwave Infrared (SWIR) and Thermal Infra-Red (TIR) regions, lithological dis- crimination and mineral potential mapping were possible from space/airborne platforms 3–5 . However, detailed un- derstanding on precise mineral composition and relative abundance of constituents within Field of View (FOV) was not possible with these data due to coarse bandwidth and poor spectral contiguity. However, when spectros- copy, radiometry and imaging techniques were bundled as imaging spectroscopy, limitations of multispectral remote sensing were overcome. The hyperspectral sensors on the other hand, are capable of acquiring images in 100–200 contiguous spectral bands. This ability to acquire laboratory-like spectra from an air/spaceborne sensor is a major breakthrough in remote sensing 6 . As a result, hyperspectral sensors provide a unique combina- tion of both spatially and spectrally contiguous images that allow precise identification of minerals 6 . Over the last two decades, mineral mapping and lithological discrimina- tion using airborne hyperspectral sensors like AVIRIS, HYDICE, DAIS, HyMAP have been extensively attemp- ted 7,8 . However, launch of NASA’s EO-1 Hyperion sen- sor with 242 spectral bands in 0.4–2.5 m range marked a new beginning in spaceborne mineral potential mapping. In this article a comprehensive review of spectroscopy of minerals and rocks, importance of field spectroscopy, and challenges in analyses of hyperspectral data for geo- logical exploration are discussed. Reflectance and emission spectroscopy of minerals and rocks Since Newton’s discovery of composite nature of white light in 1664, spectroscopy in all ranges of wavelength has been used to study properties of terrestrial and extra- terrestrial objects 9 . When light interacts with a mineral or rock, certain wavelength regions of incident light are absorbed, some are reflected, and some are transmitted depending on the chemistry and crystal structure. Absorp- tion of energy in minerals results from electronic and vibration processes of molecules 6,10,11 . The electronic processes include crystal field effects, charge transfers, conduction bands and colour centres 12,13 . The vibrational processes involving stretching, bending and rotation offer information about functional groups. Molecular vibration-related spectral absorption is charac- teristic of functional groups and is useful in identifying minerals. Absorption features related to fundamental, overtone and combination manifest in the 1–30 m region. Spectral absorption features of minerals (such as silicates, oxides, hydroxides, carbonates, sulphides, ni- trates and borates) are well established (Figures 1–3) and identification of these minerals based on spectra is now possible 14–20 . These studies on reflectance and emission spectroscopy of minerals lead to generation and archival of exhaustive spectral library.