Floods can happen unexpectedly, and often catch motorists off guard. The results can be fatal — in fact, as little as two feet of water can carry away even SUV-sized vehicles. According to the National Weather Service, more than half the deaths caused by flash floods happen when vehicles are swept away by floodwaters.

Timely, accurate information would give motorists a better chance of getting home safely, but given the vast number of road-stream intersections in Iowa, how is it possible to reliably predict when and where flash flooding will occur?

In cooperation with the Iowa Department of Transportation (DOT), researchers from the Iowa Flood Center (IFC) at the University of Iowa are working to do just that. They have initiated a two-year research project to design, implement, and evaluate an innovative flood forecasting system to help authorities better predict flooding potential in a watershed.

“Our vision is a paradigm shift in water-level observation and stream-flow modeling,” explains IIHR Research Engineer Ricardo Mantilla. With IFC Director Witold Krajewski, Mantilla has designed a hybrid flood forecasting system that combines real-time water level observations with an advanced mathematical hydrologic model. “This system gives us the best of both worlds,” Mantilla says.

Stream-stage sensors will be installed on bridges in the Squaw Creek watershed to help develop a flood forecast model for Iowa.

IFC researchers developed an affordable bridge-mounted electronic stream-stage sensor that uses sonar to measure the distance from the sensor to the water’s surface. Researchers deployed 20 of these sensors within the Squaw Creek basin in Central Iowa, upstream from its junction with the South Skunk River. The automated sensors measure stream water height and transmit the information automatically to the Iowa Flood Center, where it is available online through the Iowa Flood Information System.

The data these sensors collect are combined with a state-of-the-art mathematical model that uses detailed digital elevation models of the landscape, NEXRAD rainfall, and soils and land use data to accurately represent flow through the watershed. Sensor-collected data is used to continuously refine the hydrologic model’s reliability until it can provide dependable predictions at locations where no sensor is present.

“By the end of the two-year project, the system will provide accurate predictions of flooding potential for each and every road/stream intersection in the basin,” Mantilla says. “It will establish the basis for a customized real-time flood and flash-flood forecasting system for roads and bridges across the entire state — thus increasing public safety.”

The Iowa Flood Center is part of IIHR—Hydroscience & Engineering, a research institute based at the University of Iowa’s College of Engineering. The IFC was established in 2009 to provide accurate, state-of-the-science-based information to help Iowans better understand their flood risks. It is the nation’s first academic center devoted solely to the study of floods.