A Novel Topography-Based Approach for Real-Time Flood Inundation Mapping

Abstract

The occurrence frequency and catastrophe caused by flooding are increasing rapidly, highlighting the importance of real-time impact-based forecasting. However, traditional approaches primarily based on hydrodynamic models need large computational cost and generally fail to achieve real-time flood mapping, especially for large-scale watersheds. In this work, a novel, simple and convenient approach called Topography-based Flood Inundation Mapping (TOPFIM) is developed to achieve rapid and accurate flood mapping. TOPFIM is characterized by an adaptive river segmentation method and a dynamic inundation volume allocation approach adhering full water volume constraint. The proposed approach is applied to the upper reaches of the Le'an River basin, China, and HEC-RAS is employed as the benchmark for comparison. The results demonstrate that TOPFIM's simulation accuracy for inundation extent approaches that of hydrodynamic models, with an averaged critical success index of 0.83 and hit rate of 0.90 compared to HEC-RAS's simulation. Moreover, TOPFIM generates flood inundation mapping prediction within 10 s rather than hours required by conventional hydrodynamic models. It signifies a pivotal practical enhancement that has the potential to effectively preserve lives and protect assets in times of flood emergencies. Overall, as a simple and convenient tool, TOPFIM demonstrates its potential for real-time flood inundation mapping and risk analysis.