Regional-scale dynamic planning for distributing emergency supplies under evolving tropical cyclones

A well -designed response plan for the distribution of emergency supplies (ES) during an evolving tropical cyclone (TC) can effectively minimize socio-economic losses in coastal regions. However, the dynamic and uncertain nature of a TC is likely to result in spatio-temporally varying ES demands in the different areas within a hazardstricken region. While meteorological forecasting technologies and advanced data collection mechanisms can provide real-time information on an evolving storm and its impacts, a methodology that can systematically integrate such information into real-time ES allocation decisions has yet to be formulated. This study develops a framework to support proactive, regional -scale, multi -phase ES allocation decisions that adapt to the dynamic characteristics of a TC as it evolves in real time. The framework consists of two modules: an updating module that uses real-time information collected during the evolving storm to continuously update the projections of the ES demands in different areas within the region; and an optimization module that allocates the ES according to the updated demand projections by minimizing the ES shortages and excessive redundancies, as well as the logistics costs of the allocations. An application of the framework to ES planning in Zhejiang Province, China, during Typhoon Lekima in 2019 is presented.