The group wants to know exactly what to expect before a flash-flooding disaster strikes.
"We're talking potential life-saving actions that will be taken from these warnings in the future," said NWS meteorologist Chris Legro.
The researchers from the storms lab in Norman developed new models to show elevation, rainfall, and even direction of the water. They say it will help predict danger for people actually in the storm's path, so forecasters can issue specific warnings rather than those county or city-wide as they do now.
You can see specific neighborhoods, specific creeks and streams that are flooded," said research meteorologist J.J. Gourley who is heading up the experiment.
Forecasters from across the country are in town to test the new technology. Legro flew in from Portland to participate. He and a handful of colleagues spend hours simulating real floods and practicing sending faster and more precise warnings.
"So far I like what I'm seeing," Legro said.
The public can actually help with the experiment by downloading the M-Ping app. The app allows people to report different weather events, like a flood, and tell researchers how severe the event was near them.
"They're incredibly useful to us," Gourley explained, "we use that (information) to help validate some of our products."
Despite living in tornado alley, researchers say flooding is more deadly. And forecasters say the better models will increase warning times.
"In just the first week of the experiment here we've been able to increase lead times as much as 20 minutes, 30 minutes, 40 minutes," Legro said.
That's a major difference, when every second counts.
Another big change they hope will come from the technology is better predictions of a flood's danger: whether it's just a "nuisance flood" or is actually "catastrophic."
That way first responders know the exact intersections to watch for the highest and fastest-moving water.