Flood Susceptibility Mapping using Frequency Ratio and Shannon’s Entropy Models in the Plains of North Bihar, India

Flooding and the associated risks are increasing in the plains of north Bihar, India almost every year due to variable factors such as changes in rainfall patterns and changes in land use. For flood susceptibility mapping, an inventory map having 131 flood locations are prepared using flood annual layers obtained from Geo-Platform of Indian Space Research Organization. Out of these, 70% (92) flood locations are randomly selected as training dataset and remaining 30% (39) flood locations are selected as validation dataset. The flood conditioning factors such as elevation, land use/cover, normalized difference vegetation index, plan curvature, rainfall, slope angle, slope aspect, soil, and topographic wetness index are prepared using the geo-spatial database of ISRO. Flood susceptibility maps are developed using Shannon's entropy and frequency ratio models in ArcGIS 10.5. The validation of each model is done by calculating area under model's receiver operating characteristic curves. The area under curve for Shannon's entropy and frequency ratio are 89.93% and 87.74% respectively for the success rate and 85.29% and 85.13% respectively for the prediction rate. The validation results showed that the Shannon's entropy and frequency ratio and models have reasonable results and hence the models can be useful for flood control and mitigation measures. Citation: Sangeeta Kumari, Ninganagouda Goudar. "Flood Susceptibility Mapping using Frequency Ratio and Shannon's Entropy Models in the Plains of North Bihar, India ." Global Research and Development Journal For Engineering 6.12 (2021): 1 - 8.