This project will improve understanding of and decision support for hurricane evacuation and sheltering through a dynamic modeling approach. The hazard will be represented using a set of probabilistic scenarios that describe the range of ways a hurricane may evolve. For each hurricane scenario, storm surge, wind speed, and rainfall flooding maps will be developed for each time step. The research team will also model the dynamic decision-making of emergency managers and residents as the available information changes, and the dynamic movement of residents over the course of the event. Overlaying the results of these models will help to understand the interactions within and among the systems through space and time. The project has four main components. First, an integrated meteorological, hydrological, and storm surge modeling system that will be implemented to determine a set of probabilistic ensemble scenarios for an offshore hurricane. Second, theoretical models of resident and emergency manager temporal decision-making will be developed. Third, a multi-stage stochastic program will be developed that integrates the outputs from the first two steps to support emergency managers' hurricane evacuation and sheltering decisions so as to minimize personal risk and travel time. Fourth, the interconnected hazard, behavioral, and evacuation/shelter models will be demonstrated through case study applications in North Carolina.
This project will have broad significance because it will result in a major leap forward in understanding and improving management of evacuation and sheltering for hurricane events. Wind, storm surge, waves, rainfall, and runoff, all of which can substantially affect evacuations, will be considered simultaneously and comprehensively in this project. The project will also advance understanding of how officials and residents make decisions over time, and how the information on which their decisions change over time. It will explicitly represent the fact that emergency managers make initial decisions while a hurricane is still far offshore and large uncertainty remains, and then make subsequent decisions after a day or two has passed and more is known about the likely effects of the storm. Overall, the framework for this project will, for the first time, fully capture the three features of hurricane events that are perhaps most important for understanding evacuation and sheltering; these features are: (1)the events are dynamic over space and time, (2) involve great uncertainty, and (3) include many interactions within and among the natural, infrastructure, and human systems.The educational video game and K-8 educational outreach will teach participants about the dynamics and uncertainty in hurricane evacuation and sheltering. Post-doctoral researchers, graduate and undergraduate research assistants including women and underrepresented minorities' participation in all aspects of the project will help train the next generation of disaster professionals. Collaboration with federal and state agencies throughout the project will help ensure the findings consider practitioner's perspectives and are integrated into practice as quickly and effectively as possible.
