Remote sensing of severely degraded land: Detection of long-term land-use changes using high-resolution satellite images on the Meghalaya Plateau, northeast India

Long-term land use/land cover (LULC) assessments, based on remote-sensing data, play a leading role in gathering information on land degradation that has an impact on food security, soil quality and species habitats with all the accompanying social implications. The aim of this study was to determine the LULC changes of severely degraded land using high-resolution remote-sensing data from the last ~50 years in order to determine whether human activities have led to improvements or further land degradation. An area of 70 km2 near Cherrapunji in Northeast India, which receives 11,000 mm of rainfall annually and was deforested in historical times, was selected for this analysis. Despite severe land degradation, the population density of this area has been continuously increasing, standing at ~300 inhabitants km−2 in 2011. A visual interpretation technique, combined with field surveys, was used for LULC mapping, based on satellite images from the US Corona programme for 1965, pansharpened Indian Remote Sensing (IRS) images IRS-1D for 1998 and Google Earth for 2017. This analysis indicated the dominance of grassland (73.2%), with a small contribution of forest (17.6%), settlements (5.8%), mining (2.3%), cropland (0.7%) and water bodies (0.4%) in 2017. The LULC structure was the result of a largely degraded land contribution inherited from the past and the emergence of new driving forces behind land-use change. The LULC trends in 1965–2017 included a decrease of forest (4%) and grassland (5%), stability of cropland, and an increase of land-use classes which, in non-degraded land under a monsoonal climate, are usually of small extension, such as settlements (85%), mining (132%) and water bodies (525%). These changes reflect both, attempts to utilise existing environmental resources and to adapt to living in a degraded environment caused by growing human population. From the land degradation perspective, the trends in LULC change show two opposing directions––further land degradation mainly associated with the expansion of settlements and mining (accounting for ~79% of the total LULC changes), and land improvement and neutral change, including the recovery of natural vegetation and the construction of water bodies. The relationships between the LULC changes and direct measurements of biomass and carbon stocks of forest and grassland ecosystems indicated that their balance was negative for the Cherrapunji spur in 1965–2017. A strategy aimed at restoring the sustain