Title: Rapid decision-support tool for issuance of flood evacuation orders
Emergency managers are challenged to make effective and timely evacuation decisions prior to impending floods to reduce consequences. Traditional tools, such as hydrologic models, provide detailed information to managers about forecasted flood characteristics but do not consider how humans might behave during an evacuation event. An integrated, agent-based modeling framework is introduced to resolve human decisions and traffic dynamics during flood evacuation events. The agent-based opinion dynamics model and a large-scale agent-based transport model (MATSim) are coupled with an optimization model (SCIP) and flood inundation model (National Water Model) to develop the decision support framework. Tradeoffs between flood risks on the road and at home are minimized. The framework is applied using a machine learning-based algorithm to develop a city-specific rapid and easy-to-use decision support tool, which recommends evacuation order policies to emergency managers. Recommendations include when and where to issue suggested and/or mandatory flood evacuation orders based on the severity of a forecasted flood event and lead time to flood arrival. The tool will be installed in the platform of Urban-Flooding Open Knowledge Network (UF-OKN), a project supported by the National Science Foundation (NSF). A demonstration of model development and tool application is presented for a forecasted flood event in the City of Wilmington, NC. Efforts to apply the tool to cities across the Contiguous United States (CONUS) are discussed.
Keywords: Flood evacuation, decision-support tool, agent-based modeling, optimization, MATSim
Michael recently graduated from the WRES program at UIUC after completion of a master’s degree program within Dr. Ximing Cai’s research group. He now works at the Construction Engineering Research Laboratory (CERL) in Champaign where he is a part of the Water Use Resilience, Security, and Technology team for the U.S. Army Corps of Engineers (USACE) Engineer Research and Development Center (ERDC). From his time as a consulting engineer and researcher with Dr. Cai, Michael became interested in natural hazards, modeling of complex decisions, and exploration of sustainable land-use practices.