BATON ROUGE, La. (BRPROUD) – For years, LSU has been part of developing a system that helps simulate storm surges and flooding during weather events. Now artificial intelligence is being researched as a way to improve the system, which is used by government agencies to determine when floodgates should be shut and make decisions on emergency orders.

LSU helped create the Coastal Emergency Risk Assessment (CERA) system over the last several years. The map takes in data from a number of partners as they forecast incoming storms. It then turns it into graphs and shows on the map where storm surge is expected and how much, with a certain amount of accuracy.

“As you can imagine. These simulation results are not simple. It’s essentially millions of numbers, if you want, that need to be presented in a way that you and I can understand it,” said LSU research professor Hartmut Kaiser. “And really see this house gets flooded and this house won’t get flooded.”

The CERA tool helps emergency agencies and the general public see where stormwater potentially could go using maps that can show your own neighborhood. Compiling all the data for the simulation is complicated and the system is always being improved to be more accurate. In the case of Hurricane Ida, a storm that wobbled on where it would make landfall, the CERA system could show the different possibilities for storm surge with the changing landfall site.

New research has begun on the system using artificial intelligence to help speed up the process of creating and validating the simulations by using historical data.

“Use that data over time to predict the error during the forecasting period so that we can correct the simulation data based on that comparison using real-time data during the visualization,” Kaiser said.

Agencies like the Coastal Protection and Restoration Authority use the CERA system to help communicate with levee districts on when floodgates should be closed. 

“We know the elevation of the surrounding area, we know what the storm surge is going to do, we have our operations manuals as to when to close these gates,” said CPRA Operations Chief Ignacio Harrouch. “We see a trigger, you see the hydrographs, and you call them and say we need to close this gate.”

The CERA model is just one of the many tools that helps them get a good idea of what is coming as the storm advances and how different districts need to prepare.

“We’re pretty confident as to what’s going to happen. 10-15 years ago we didn’t have all of this. We were trying so we were at mercy,” Harrouch said. “Nowadays we can prepare way ahead of time with the information we currently have.”

The information from the simulations are sent to elected officials and emergency management teams to aid them in determining evacuation orders, road closures, and more. 

The research team across multiple universities, with a grant from the Department of Energy, looks to improve the CERA system with AI technology for future storm preparations.