A new two-stage emergency material distribution framework for urban rainstorm and flood disasters to promote the SDGs

Abstract

Under the coupled effects of climate change and urbanization, urban rainstorm and flood has become more frequent and intense. The effect of the flood disasters emergency response has a direct impact on the success or failure of the rescue. Emergency material distribution is one of the most critical steps in the emergency response phase and is crucial to achieve the Sustainable Development Goals (SDGs). This study proposes a two-stage emergency material distribution framework for urban rainstorm and flood disasters. The entropy weight method and technique for order preference by similarity to ideal solution (EWM-TOPSIS) was adopted to incorporate the difference in relief-demand urgency into the emergency material distribution principle. The mathematical model was established by minimizing emergency travel time, inequity and social costs (logistic costs and deprivation costs) as objective functions. The gamultiobj algorithm was used to solve the model. The proposed framework is further applied in Xi'an city of China. The results showed that the affected sites were categorized into disaster risk levels I, II and III. The Pareto solution set was obtained by optimizing each conflicting objective. The emergency material distribution plan was selected from the Pareto solution set based on the actual situation. In the plan, the flood disasters emergency travel time totaled 6.48 h, the inequity totaled 3.69, and the total social costs reached 226,584 CNY. The proposed framework is more consistent with the needs of the actual rescue by comparing the four models generated plans. Decision-makers should not only consider minimizing dispatch time and social costs, but also the trade-off between equity and prioritization of distribution. This study helps to provide new ideas for sustainable urban development and emergency management. Urban rainstorm and flood disasters change with the changing environment dynamically, and it is difficult to collect dynamic disaster information on the affected sites. Thus our further study will focus on improving the prediction accuracy of dynamic urban rainstorm and flood disasters.