Assessing the Importance of Static and Dynamic Causative Factors on Erosion Potentiality Using SWAT, EBF with Uncertainty and Plausibility, Logistic Regression and Novel Ensemble Model in a Sub-tropical Environment

Abstract

The sub-tropical countries like India experience large-scale land degradation due to erosion of the surface soil. So, there is a direct impact of monsoon climate in this region; large-scale water-induced erosion has gradually increased due to an extreme precipitation event. There is an unbalanced situation that happens between regolith formation and rate of surface soil erosion. The amount of agricultural production is decreasing day by day due to the degradation of topsoil. The gap between demand for the crops and the rate of production is thus rapidly increasing. Evaluation of possible erosion zones is needed to identify vulnerable areas for the implementation of suitable remedies. Estimating the morphotectonic parameters and their effect on potential for erosion is important for determining the most possible erosion zones. Here, most of the morphometric and tectonic parameters are considered as static and dynamic causative factors. Due to the diversified hydro-geomorphic association, in recent times the rate of soil erosion in the Kangsabati River Basin is very high. Here, the SWAT modelling tool was considered for the estimation of the morphometric parameters in a systematic way. The evidential belief function (EBF) with uncertainty and plausibility, logistic regression (LR) and ensemble EBF–LR were then used to estimate this region’s erosion potentiality. For estimating the erosion potentiality, the location of gully and non-gully points is considered and that is randomly divided into 70/30 ratio. It can be said from this analysis that most of the watershed of this river basin, the erosion potentiality rate, is moderate to very high. Thus, these watersheds need to be carefully analysed by implementing certain structural and non-structural measures to reduce the potential for erosion.