Journal of Soil and Water Conservation

Remote sensing based identification of salt-affected soils subject to differences in spectral signatures using multispectral and hyperspectral data is the most common approach. Based on the literature, an attempt has been made to review the appropriate season, sensors and techniques used to map and monitor the three classes of salt-affected soils (saline, sodic and saline-sodic soils). The choice of season of satellite data for mapping of salt-affected soils depends on objective, crop rotation and rainfall pattern in an area. Both saline and saline-sodic soils show white reflectance but field survey and analysis of soils is required for differentiation of these soils. The visible and near-infra red bands are useful for separating saline and saline-sodic soils from non-salt-affected soils. In general, the surface of sodic soils become smooth due to hard crust and dark coloured due to dispersion of organic matter, and these indicators may be used to separate sodic soils from saline and saline-sodic soils when satellite data of appropriate season is taken. Generally thermal band is useful for separating sodic soils from saline and saline-sodic soils. The remote detection of salt-affected soils using various techniques (like principle component analysis, vegetation indices, Bayesian and neural networks, Support Vector Machine, Minimum Distance and Maximum Likelihood) has been reviewed and the conditions have been highlighted where the different approaches are likely to be most effective. The recent developments in satellite remote sensing support early detection of salt-affected soils, however, accurately detection of small patches of these soils is a challenge.