Monitoring of agricultural soil degradation by remote-sensing methods: a review

Agricultural land degradation is a global problem that severely hampers the production of food needed to sustain the growing world population. Mapping of soil degradation by remote sensing is instrumental for understanding the spatial extent and rate of this problem. Methods aimed at detecting soil loss, soil drying, and soil-quality deterioration have been demonstrated in numerous studies utilizing passive and active remote sensors. This review presents a short description of each form of soil degradation, including data regarding known extents and rates, and then reviews the methods with respect to direct and indirect modelling approaches. Two types of obstacles to achieving wide regional detection of soil degradation are revealed. The first concerns the complex and non-unique relationships between remote-sensing indicators and different soil properties, such as roughness, soil moisture (SM), soil salinity, and organic matter content. The second concerns the difficulties involved in acquiring data on subsurface soil properties. In view of these difficulties, we recommend expanding the use of phenomenological models capable of estimating soil-degradation potential according to combinations of environmental conditions, thus enabling remote-sensing efforts to be focused on local areas where the environmental threat is highest. The second avenue for improving the contribution of remote sensing on a wide regional scale involves the application of integrative methods, such as those based on hyperspectral, multisensory, and multitemporal approaches, as well as those that incorporate environmental information (such as topography, soil types, and precipitation).