Flood disaster risk assessment in the east of Sichuan, China

It is important to assess the flood disaster risk in eastern Sichuan, China, as it is a densely populated and flood-prone area. In this paper, the Dempster-Shafer (DS) evidence theory is applied to assess the flood hazard risk because of the theory's abilities in multisource information fusing and uncertainty processing. The Variable Fuzzy Sets (VFS) model establishes the basic probability function in the DS model. This coupled DS_VFS model is then employed to assess flood risk levels. The precipitation record, elevation, topographic slope, socioeconomic data and historical flood frequency are used in the analysis process in Sichuan. As an example, the belief function, relief function and uncertainty of the high (H) risk level in Chengdu (a sub-region in our study area) are determined as 0.857, 0.858 and 0.001, respectively, which suggests a high flood risk. As a result, the whole area can be divided into several blocks, and the risk levels were higher in the central and western areas, but lower in the east. Especially, Chengdu and Nanchong got the highest risk levels due to dangerous disaster-inducing factors and a fragile social economy system. In order to verify the DS_VFS's result, the traditional risk matrix analysis method was used to determine the hazard risk levels, and validation showed that the evaluation results were reasonable and reliable. In addition, more calculation methods and decision fusion theories will be introduced into flood risk assessment in future research.