Predicting Erosion Potential Zones using CA-ANN and Multi-Criteria Analysis for the Bhavani Watershed, India

Abstract

Watershed prioritization has gained importance for locating critical areas of erosion and for taking up proactive steps in watershed management. The present study attempts to forecast land use/land cover change and prioritize the sub-watersheds of Bhavani watershed, Tamil Nadu based on morphometric, land use/land cover and slope coupled with the multicriteria analytical (MCA) framework using remote sensing and GIS. The land use/land cover is predicted for 2030 using an artificial neural network-based cellular automata simulation model, and the sub-watersheds sensitive to erosion were identified and compared for the years 2020 and 2030. Parameters having direct and inverse effects on erosion risk were used as the criteria in the ranking process. Using geospatial data and GIS, a set of eleven morphometric parameters, eight LULC classes, and five slope classes was analysed using a compound value weight calculator for the 26 sub-watersheds. The study demonstrates significant land use changes on cropland, built-up land, wasteland, fallow land and forest land between 2020 and 2030. The entire watershed was divided into four zones based on the prioritization approach and were categorized into very high, high, medium, and low priority zones. The results showed that SW-19, SW-22 and SW-26 were categorized under very high priority zones for the years 2020 and 2030. Besides, the SW-07 was added under very high priority zone for the year 2030.