When commentators talk about smart cities, they usually refer to buildings that regulate their own energy use, autonomous transportation systems or apps that provide real-time information about train and bus service. Jonathan Goodall, an associate professor of civil and environmental engineering in the University of Virginia's School of Engineering & Applied Science, brings an innovative perspective that could turn the tide in low-lying coastal cities from Virginia Beach to New York, where flooding caused by sea-level rise and more intense storms has become a pressing problem. He researches intelligent systems that could help localities better manage their stormwater infrastructure.
“As climate change reshapes our environment, managing stormwater will increasingly be a challenge in areas that are home to a large portion of the world’s population,” said Goodall, who is assistant director of UVA Engineering’s newly created Link Lab for cyber-physical systems.
Goodall's approach has attracted the attention of the National Science Foundation (NSF), which will support the work with more than $4 million in new funding through two grants. The first is a $2.5 million award, part of the NSF’s Critical Resilient Interdependent Infrastructure Systems and Processes solicitation, to a team of faculty members who will design a decision-support system to assist infrastructure managers reduce flooding risk in Norfolk, Va. In addition to Goodall, the team includes UVA Civil & Environmental Engineering Assistant Professor Donna Chen; Science, Technology & Society Professor Michael Gorman; and Associate Professor Kamin Whitehouse and Assistant Professor Madhur Behl, both from the Department of Computer Science.
The NSF awarded a second grant to a multi-institutional group that includes Goodall and Civil & Environmental Engineering Associate Professor Teresa Culver. Offered through the agency’s Smart and Connected Communities solicitation, the $1.9 million grant is for research on overcoming the social and technical barriers to the broad adoption of smart stormwater systems.
Dynamic Stormwater Management
Stormwater infrastructure is, for the most part, static. Localities build retention ponds and other facilities designed to minimize the impact of impervious urban surfaces on water quality as well as quantity. The addition of sensors and automated valves and pumps to these facilities are beginning to make these defenses more responsive and effective. For instance, in the face of a storm, engineers can lower the water level in ponds to increase their capacity.
“The systems that can provide real-time sensing and control are fairly mature in many infrastructure systems,” Goodall said. “They are starting to be introduced to stormwater infrastructure.”
The University of Virginia has installed a model system just yards from the Department Civil & Environmental Engineering. The drainage pipes in this small bioretention facility can be adjusted with computer-controlled valves. At the same time, the facility is designed to improve water quality. Plants growing in the facility consume nitrogen present in the runoff before the water percolates through the soil and leaves through underground drainage pipes.
“Students are developing algorithms to maximize the pond’s ability to purify water while handling the quantity of water produced by storms,” Goodall said. “With smart technology, you can make a treatment facility more effective than it would be otherwise.”
Building Models
Smart facilities are only one part of the solution. To make prudent decisions about using them, engineers must have models that enable them to predict in detail the flooding produced by storms and tides. This is Goodall’s area of expertise.
“Right now, these forecasts are primarily based on rainfall, which is a pretty good indicator of flooding,” Goodall said. “But if you really knew the landscape and had detailed information about flooding caused by past storms, you would have a more accurate basis to close roads or actively manage stormwater infrastructure.” You would also have better rationale for making other decisions like closing roads.
There are a number of challenges Goodall faces. One is the shortage of data needed to produce reliable algorithms. A solution would be to place more rain gauges in a watershed — or to develop indicators that would serve as proxies for rain gauges. Subscribers to sites like WeatherUnderground.com can register their own rain gauges with the website, and Goodall’s team is making use of this information.
He and his colleagues are also looking to incorporate new sources of flooding data. For instance, the City of Norfolk has introduced an app that allows citizens to report when roads are flooded. Goodall is collaborating with researchers from Old Dominion University to determine if machine vision could be used with traffic cameras to identify flooded intersections.
Goodall and his colleagues are also incorporating geographic information system data layers in their models. They include elevations as well as maps of roads and storm sewer networks.
“We are gathering everything we can from local governments and other sources,” Goodall said. “We are drawing a wide net.”
A Broader Vision
As Goodall realizes, knowing how to operate a smart stormwater system and implementing it effectively to increase a community’s resiliency are two different things. Engineers have to consider the social barriers to adoption. Residents, for instance, might be concerned about the risks posed by active management techniques or object to changing water levels that might result. To address these issues, Goodall is working with urban planners, architects and social scientists. UVA’s newly created Pan-University Environmental Resilience Institute, for which Goodall has served on the steering committee, will be critical to strengthening these collaborations.
Goodall also recognizes that effective stormwater management will require a menu of approaches in addition to smart infrastructure, one that might include restoration of wetlands and even physical barriers as well as coordination among a variety of jurisdictions. Network-level control is essential, he argues, which requires coordination and compromise among a multitude of private and public entities.
“We need to think holistically in terms of the large portfolio of things we can do,” Goodall said. “And then we need to develop a framework that would allow these entities to collaborate as part of a broad system.”
The problem is challenging, but Goodall believes communities can get ahead of it with collaborative and innovative approaches.