Prediction of highly flood prone areas by GIS based heuristic and statistical model in a monsoon dominated region of Bengal Basin

Flood has become one of the major environmental disasters in the world. A flood susceptibility assessment for the River Dwarkeswar was performed in this article, and the outcome was compared with the Koiya River of Bengal Basin, India. Fourteen flood conditioning parameters such as normalized differential vegetation index (NDVI), rainfall, stream power index (SPI), land use and land cover (LULC), topographic wetness index (TWI), geology, soil, slope, elevation, drainage density, plan curvature, profile curvature, aspect and distance from river etc. were identified for both the basins to determine the susceptibility to flooding. Four models such as Analytical Hierarchy Processes (AHP), Knowledge Driven (KD), Fuzzy Logic (FL) and Logistic Regression were applied. Area under curve (AUC) analysis shown that in Dwarkeswar River LR (AUC = 0.916) is very much successful compared to AHP (AUC = 0.869), KD (AUC = 0.841) and FL (AUC = 0.893) whereas Koiya River of Bengal Basin shows that LR (AUC = 0.902) is more accurate than FL (AUC = 0.879), AHP (AUC = 0.861) and KD (AUC = 0.828). The outcome of this study can be used by the planners and policy makers to implement the management measures with a view to the local environment.