Researchers use artificial intelligence, machine learning technology to predict coastal flooding

Katy Nelson, News Reporter

Editor’s note: This article was originally published in the Jan. 25, 2022 flipbook.

A UT researcher, in collaboration with other schools and the Department of Energy, is using artificial intelligence to develop better strategies for flood predictions and preparedness.

As part of the collaboration, University researcher Clint Dawson utilizes artificial intelligence and modeling to analyze weather data. Dawson said AI and machine learning technology could help predict future storms and flooding as a result of rising sea levels and irregular weather patterns. Their research, which began in September, is set to finish Aug. 2024.

“We’re trying to produce a computational tool,” Dawson, an aerospace engineering and engineering mechanics department chair said. “So that stakeholders and emergency management agencies can use (it) to study things like flood mitigation, flood risk, impacts of future climate (and) the impact of sea-level rise.”

The researchers enter weather data into a computer algorithm to simulate future forecasts. Researcher Joannes Westerink said their work is crucial because the mathematical formulas that have been used to predict weather events in the past lack computer technology.

“The equations that represent (the) physical system of the ocean tides that were used for storm surge and a lot of hydrology were developed 245 years ago,” said Westerink, a computational sciences and engineering professor at the University of Notre Dame. 

The AI technology allows researchers to simulate future weather events, which the mathematical formulas are incapable of on their own.

The researchers use data from the DOE’s Energy Exascale Earth System Model on coastal areas to run the AI simulations and develop improved flood mitigation and prediction. Westerink said the model helps to understand how a changing climate can impact coastlines and flooding risks.

The group’s collaboration is funded by the Department of Energy’s Office of Advanced Scientific Computing Research and will receive a total of $5 million over three years. Dawson said their innovative flooding mitigation methods are already underway in areas of Texas, such as a potential levee system in the Houston Ship Channel. 

“We want to improve our capabilities to predict flooding events caused by a storm surge,” said Hartmut Kaiser, a senior scientist at the center of computation and technology at Louisiana State University. “The storm surge impacts mostly the immediate coastal region but rainfall impacts much larger regions far inside the land, and we hope to be able to better predict these flooding impacts.”