Evaluating Human Casualty Estimation Models for River Floods on a Regional and Global Scale (Case Study: Can Watershed)

Flood is one of the most important types of natural disasters that lead to serious damage, especially in Iran. Despite the high number of human casualties in Iran, this phenomenon has not been sufficiently studied. A review of the relevant research conducted so far shows that the number of relationships is low and at the same time their accuracy is questionable. In this study, the developed regional and global models have been evaluated to determine human casualties in the catchment area of Kan River located in Tehran. In this research, HEC-HMS has been used for hydrological simulation and HEC-RAS two-dimensional model has been used for flood simulation and estimation of parameters such as velocity, depth, and rate of flood rise in an unstable state. The spatial distribution of all parameters of both regional and global models according to the structure of the models was obtained in the GIS environment. Finally, the short-term risk of flood losses in the study basin was calculated using the results of the regional model. The regional model which was developed based on the hydraulic and morphological parameters of the basin and the global model which was designed for mortality due to dam failure were able to simulate historical events well, however, the regional model in all high-risk to low-risk areas has better results. The global model, which is a parametric model, was found to be acceptable for estimating model losses in the range of low-risk and predicting instantaneous flood mortality, by changing the average distribution parameter by twice the standard deviation. The use of both regional and global calibrated models is recommended in risk management and instantaneous floods crisis.