Daniel Wright hopes to one day transform the way we analyze and estimate flood risk in the United States and around the world.
That’s a lofty ambition that goes well beyond a single research proposal. But the assistant professor of civil and environmental engineering at the University of Wisconsin-Madison sees the work he’ll undertake as the recipient of a National Science Foundation CAREER award as the next step on his longer quest.
Researchers have traditionally studied floods from two disparate perspectives: the physical point of view, which examines how rainfall interacts with the land and rivers that comprise watersheds; and the statistical approach, which attempts to define the relationship between flood severity and likelihood (known as flood frequency analysis).
“There’s a lot of middle ground between those two approaches that hasn’t really been very well explored,” Wright says.
He hopes to bridge that gap through his NSF CAREER grant, which provides more than $500,000 over five years. To do so, he plans to use data analysis and modeling that starts, fundamentally, with what he calls the ingredients of floods: rainfall, land cover and use, and soil moisture.
“It’s pretty ambiguous right now in terms of what the future holds for flood risk in a changing climate,” he says. “The reason is that big floods are so rare that it’s hard to infer directly from really infrequent observations how the world is changing, and specifically how flood risk is changing. I think we stand a much better chance of doing that if we focus on how the individual ingredients are changing.”
To better understand the first and foremost ingredient—rainfall—Wright will analyze records of every rainstorm across the United States over the past 70 years, thanks to an algorithm created by collaborators at the University of California, Irvine. Wright says typical examinations of rainfall trends over time have relied upon data from rain gauges.
“We actually get to look at the size and shape and motion of these storms, as opposed to just looking at what the rain gauge is telling us,” he says.
He’ll also build upon work he did as a doctoral student at Princeton University and NASA’s Goddard Space Flight Center, where he developed software called RainyDay. It allows users to develop thousands of hypothetical rainfall scenarios in order to run flood simulations for specific watersheds with different land use and soil moisture conditions.
Wright will use RainyDay in combination with observations from the National Weather Service’s NEXRAD network of weather radars, which provide extremely detailed pictures of rainstorms and how they evolve over time. Ultimately, global rainfall observations from weather satellites will allow lessons learned in this research to be applied to understand flood risks in other parts of the world.
“We’ve been showing that this approach can give us a richer picture of what this ‘frequency versus severity of flooding’ is than some of the more conventional statistical approaches can,” he says. “And it’s doing so because it’s starting from really representing the physics.”
The outreach components of Wright’s CAREER project also lean heavily on technology, including using an “augmented reality sandbox” to illustrate concepts like topography and runoff for K-12 audiences; working with graduate students to develop web-based tutorial apps on key hydrology and statistics concepts; and creating educational modules based on the physics of the video game Minecraft.