How Argonne project could help communities better prepare for extreme flooding

Researchers at Argonne National Laboratory in DuPage County are using supercomputers to develop a dataset that provides hyperlocal flood risk estimates.

Scientists plan to apply the dataset, which shows increased risk from climate change at the neighborhood level, across the country. While it's well-documented that global warming increases the frequency of extreme weather events such as severe inland flooding, researchers hope the development at Argonne will help utilities, emergency response managers and governments better prepare at the local level.

“The central story behind our research is we want to help by providing datasets to communities — that could be critical installations, utilities, like the power sector — with projections of climate data,” Rao Kotamarthi, science director at the lab's Center for Climate Resilience and Decision Science, said. “We want to calculate the climate impact, the end result and the risks you will have because of the changing nature of these hazards.”

Funded by the New York Power Authority, a state-owned power utility, Argonne scientists recently used the dataset to look specifically at the Northeastern U.S.

Using the lab's supercomputers, researchers achieved a detailed, high-resolution simulation of the Northeast, breaking the region down into cells as small as 200 meters by 200 meters. With cells roughly the size of a tennis court, this level of analysis can be compared to a typical global scale model, which uses grids as large as 100 kilometers by 100 kilometers.

The study found that the amount of rain New York City will receive will increase by nearly 10%. It also found that the general stream flow rate in the region's rivers will increase, and the number of places that will see flooding in the Northeast will also increase.

Kotamarthi and his team are currently working on applying the dataset to the lower 48 states, which will be added to Argonne's public mapping tool called the Climate Risk & Resilience Portal, or ClimRR.

The researchers are currently working with the Federal Emergency Management Agency on how to best present the data so that emergency managers and community leaders can use it effectively.

Kotamarthi added that these calculations take time. For instance, the researchers have divided the 48 states into six domains. It takes three to six weeks for the supercomputers to run a computation for one of those domains.

“It's pretty computationally intensive, and of course, when you do a calculation, you have to check to make sure that it is correct. Once in a while you goof up and you have to go back and do this again,” Kotamarthi said. “We're constantly checking our model results to make sure that they are correct, so it requires both big computers and people who know what they're doing to do this.”

While the local flooding data has not yet been added, the ClimRR tool is currently available and provides national visualizations, local climate projections and a data catalog.

Jenny Whidden is a climate change and environment writer working with the Daily Herald through a partnership with Report For America supported by The Nature Conservancy. To help support her work with a tax-deductible donation, see .