Researchers have developed a mathematical model to predict which neighborhoods should be prioritized for search and rescue operations in the wake of a hurricane, with the goal of expediting recovery operations by the Coast Guard or other responders.

“In the first 48 hours of a major disaster like a hurricane, responders show up from all over the country to help and are often operating in an information vacuum,” says Brandon McConnell, co-author of a paper on the work and an associate research professor in North Carolina State University’s Edward P. Fitts Department of Industrial and Systems Engineering.

“Our goal with this work was to create a model that predicts where the highest rescue needs will be in order to inform operational rescue planning for the first 48–72 hours after a hurricane.”

“Specifically, we developed a predictive modeling framework to identify census tracts where residents are most likely to require rescue,” says Ben Rachunok, corresponding author of the paper and an assistant professor in NC State’s Fitts Department. “Responders will ultimately look in every area, but which areas are most likely to have people who require rescuing? If we can predict that, we can prioritize search efforts in those areas.”

For this work, the researchers drew on the available research into which factors make people more vulnerable during hurricanes, such as physical disabilities, fewer financial resources, and so on. The researchers then created a mathematical model that draws on U.S. Census data to identify which areas had populations who were most likely to have trouble leaving the area in advance of a hurricane, and draws on National Flood Insurance Program data to identify areas at greater risk of flooding.

“This work was also informed by the first-hand experiences of our first author, Patrick Leavitt, who is an active-duty Coast Guard officer,” says Rachunok. “His practical experience with emergency response operations definitely played a role in how we approached the work.”

To demonstrate the functionality of the framework, the researchers did a case study focusing on Hurricane Harvey, a Category 4 storm that struck Texas in 2017, causing catastrophic flooding in the Houston metropolitan area.

For the case study, the researchers plugged regional Census data and National Flood Insurance Program data into their model to identify areas most likely to have residents who would be trapped by floodwaters. They then compared these high-priority areas to publicly available data on where rescues actually took place in the wake of Hurricane Harvey.

“Our framework did pretty well – it should be useful for responders in practice,” says Rachunok. “It’s not perfect, but even this version would be helpful – and we can put in the work to make it even better.”

“Being able to achieve these results with this initial version suggests we’re optimistic about its utility if we fine-tune the tool – particularly in instances where we have access to better data,” says McConnell. “And the model does not take long to run; someone could run the model as responders are deploying, giving them information they can use to prioritize their efforts as soon as they’re on the ground.”

“This research is not only beneficial in the response phase of a disaster, but in the planning phases as well, assisting emergency managers in the development of response plans and exercise development,” says Leavitt.

“We’re open to working with emergency management and disaster response leaders to identify ways in which we could improve the model itself, and to inform how we can make this tool more user-friendly for practical use,” says Rachunok.

The paper, “Anticipating Household Rescue Demand in Hurricanes Using Socio-Demographic Data and Machine Learning,” is published open access in the International Journal of Disaster Risk Reduction. Leavitt, the first author, began work on this project while a graduate student at NC State. The paper was co-authored by Fred Livingston, an associate teaching professor in NC State’s Fitts Department.