Hyperspectral remote sensing: opportunities, status and challenges for rapid soil assessment in India

Abstract

Rapid and reliable assessment of soil characteristics is an important step in agricultural and natural resource management. Over the last few decades, diffuse reflec-tance spectroscopy (DRS) has emerged as a new tool to obtain both qualitative and quantitative informa-tion on soil in a non-invasive manner. The DRS ap-proach is attractive because both the proximal and remote mode of measurements may be adopted to es-timate multiple attributes of soil such as physical and chemical soil properties and nutrient contents from a single reflectance spectrum. Hyperspectral imaging cameras onboard remote sensing platforms are already providing hundreds of narrow, contiguous bands of reflectance values and the technology is becoming popular as the hyperspectral remote sensing (HRS) approach. The main objective of this review is to summarize the preparedness and opportunities for using the HRS approach for soil assessment in India. Detailed literature review suggests that the HRS approach requires large spectral databases and robust spectral algorithms in addition to the capability to in-terpret HRS images. Over the last decade, few efforts have been made to create spectral libraries for Indian soils. However, most of these libraries are very small, precluding the development of robust spectral algo-rithms. Specifically, the availability of HRS data and robust retrieval algorithms for soil properties from HRS data through unmixing procedures require spe-cial attention. With several global initiatives to make HRS data available, coordinated efforts are needed in India to build comprehensive spectral libraries, algo-rithms and create trained human resources to take full advantage of this emerging technology. Specifi-cally, a dedicated spaceborne mission will provide quality hyperspectral data for the effective application of HRS for soil assessment in India. Keywords: