A new artificial intelligence model developed by a team at Florida International University (FIU) aims to help water managers in Florida respond more quickly and effectively to flood threats. The system of canals that runs through the state is crucial for flood control, especially during hurricanes, but predicting flooding remains a challenge due to rapidly changing weather conditions.
Traditional models used for flood prediction are detailed and precise but require significant computational resources and can take up to an hour to deliver results. This delay can be problematic when storms are moving quickly.
The FIU research team has created an AI-based tool that delivers near-instant simulations of possible flood scenarios and recommends strategies for response. Their findings were published in the Journal of Water Resources Planning and Management.
“Accuracy is obviously very important to us, because overestimation of water stages can cause false alarms and panic while underestimation can result in unexpected and dangerous flooding events,” said Giri Narasimhan, professor at FIU’s Knight Foundation School of Computing and Information Sciences.
“We were able to create a tool that provides water managers with the information to either eliminate a flood event or drastically reduce it.”
Jimeng Shi, who recently graduated from FIU, led the development of the model as part of his doctoral research. The AI was trained using nearly ten years’ worth of environmental data collected by the South Florida Water Management District. It incorporated information from major storms such as Hurricane Irma (2017), Hurricane Sandy (2012), and Tropical Storm Isaias (2020). The model was tested on the Miami River, which flows through downtown Miami into Biscayne Bay.
The researchers view this project as part of their broader efforts to make AI useful in critical real-world situations where rapid decisions are necessary.
Study co-author Jayantha Obeysekera, director of FIU’s Sea Level Solutions Center and former chief modeler at the South Florida Water Management District, said the technology could also support long-term planning decisions for infrastructure investments. “The model also holds a lot of potential as a tool to help agencies make longer-term decisions,” he said. “It could guide 20- or 30-year infrastructure investments, such as whether to build new pumps, reservoirs or levees by screening potential solutions efficiently.”
