Modelling soil erosion risk of a tropical plateau basin to identify priority areas for conservation

Abstract

Soil erosion induced by physical and anthropogenic activities needs serious concern to deal with. In recent days, watersheds have been facing acute soil loss because of deforestation, poor land use management and unscientific agricultural practices. The present paper highlights the goal to decipher soil erosion-susceptible zones in a rain-fed river basin of India having a complex topography comprising the extension of Chotonagpur plateau and parts of Bengal basin. Fuzzy logic algorithmbased analytical hierarchy process (FAHP) with remote sensing and geographical information system (GIS) were used to execute the objective. The soil erosion-susceptible zone (SESZ) was identified assimilating the 14 geo-environmental, hydrometeorological and anthropogenic conditioning factors such as slope, drainage density, relative relief, stream power index, topographic wetness index, runoff, soil texture, land use and land cover, rainfall erosivity factor, distance from road, distance from settlement, distance from river, geomorphology and soil bareness index based on which sub-basin prioritization for sustainable soil conservation was made. The soil erosion-susceptible zone was classified into five classes, namely very low soil erosion zone (13.20%), low soil erosion zone (25.75%), moderate soil erosion zone (27.62%), high soil erosion zone (21.90%) and very high soil erosion zone (11.53%). Receiver operating characteristics (ROC) curve was generated where area under the curve (AUC) came as 0.82 and a correspondence analysis between the SESZ and the first three ranked factors was also made which together proves the accuracy of the model. KeywordsFuzzy AHP • Soil erosion-susceptible zone (SESZ) • Receiver operating characteristics curve (ROC) • Area under the curve (AUC)