An overview and trends of flood detection, hazard, vulnerability and risk assessment

Flooding is one of the most frequent natural disasters that result in human casualties, economic disruption, and material ruin. Flash floods are often regarded as devastating natural disasters. This research aims to examine existing literature on flood detection, flood hazard, susceptibility, vulnerability, and risk assessment to draw conclusions and recommendations by reviewing 85 recent articles related to this topic. The risk of floods results from the interplay of two factors: exposure to potential danger and exposure to potential flood damage. Assessment of flood risk requires careful examination of several morphometric characteristics related to river basins. However, this article discusses how Remote Sensing (RS) and Geographic Information System (GIS) techniques were used to identify and forecast flash floods. Compared to conventional methods and numerical methods, deep learning models of flood detection, hazard and vulnerability exhibited improved accuracy and speed, respectively. The results show that Deep Learning models are often more accurate because of the applied inductive biases to process the spatial features of flood actions extra efficiently. Global flood disaster risk is rising due to demographic, socioeconomic, unplanned settlement, environmental degradation, pressure on natural resources, and climatic factors. Decisions, policies, and management of floods all benefit from thorough risk analysis. This analysis aims to aid modellers and practitioners to choose the best flood hazard assessment approach for a given case study.