Integrated morphometric and machine learning-based flood risk assessment in the Rangit river sub-watersheds, Eastern Himalayas

This research article offers an in-depth investigation of the morphometry of the Rangit River Basin to clarify its hydrological behavior and characteristics. Morphometric analysis has been undertaken to prioritize six sub-watersheds of the Rangit River Basin, a tributary of the Teesta River. Heavy precipitation increases the likelihood of overflowing the dendritic drainage system in the lower mountain region. Sub-watersheds have been created via ArcMap 10.3 software using a Digital Elevation Model (Aster DEM) having a spatial resolution of 30 m. The total Rangit River Basin was split into six sub-watersheds labeled as SW1 through SW6. Fourteen distinct morphometric criteria pertaining to linear, relief, and areal characteristics have been developed and applied to rank and prioritize sub-watersheds for necessary examination. The sub-watersheds have been classified into three categories depending on their maximum and smallest prioritized scores for conservation and flood control. After analysing all of the morphometric criteria and evaluating the priority zones, SW1 is classified as high priority, followed by SW4 and SW5, while SW2, SW3, and SW6 are classified as low priority. A flood risk zone map of this river basin was created using 24 variables, as well as other more significant parameters: elevation, slope, and rainfall, LULC, NDVI etc. using Weka’s Random Forest machine learning algorithm. The flood zonation map and the sub-watershed-based prioritisation illustrate the areas and watersheds that are prone to floods. This research may assist future initiatives for governmental and non-governmental organizations and other scholars. Morphometric and machine learning techniques were used to determine flood risk zones in the Rangit River Basin. SW1 is classified as high priority, followed by SW4, SW5, SW2, SW3, and SW6. The basin has a high risk of flooding. By pinpointing high-risk areas for effective disaster response, this study supports efforts to mitigate flooding.