Presentations, small groups discussions, and collaborative activities to formulate knowledge graphs at the workshop on April 21 and 22, 2026 at the Institute of Behavioral Science.
Flooding is one of the most ubiquitous and costly hazards across the United States. To understand flood risks and protect communities, robust scientific data is critical. Yet existing flood data is often siloed within disciplines and organizations, making it difficult for researchers and practitioners to use it in ways that can readily contribute to risk reduction.
That’s why the Natural Hazards Center—in partnership with the National Oceanic and Atmospheric Administration (NOAA) Weather Program Office—recently held a Workshop on Advancing Social and Physical Science Data to Reduce Flood Losses.
"Whether you’re in Maine, Alaska, or the Northern Mariana Islands, flooding is going to be one of the hazards communities encounter," said NOAA Social Science Program Officer Alison Krepp in her opening remarks. She them emphasized a key question motivating the workshop itself: “How can we integrate the flood datasets that have been developed in research and practice over the years and bring them together for the public good?”
The Workshop gathered 46 data experts including social and physical scientists, engineers, urban planners, and other professionals from a wide variety of academic institutions, federal agencies, and non-profit organizations. The two-day event allowed participants to engage in discussions and hands-on learning from each other and experts, as well as pathways to improve data discovery, integration, and stewardship in the broader hazards community.
“It was so inspiring to be in the room with talented scholars and practitioners who each look at a different angle of flooding, but have a shared goal of reducing harm,” said Jessica Austin, Natural Hazards Center Research Associate.
Flood data—from spatial hazard maps and hydrologic models that show where flooding occurs to socioeconomic vulnerability data that show who will be impacted—is key to understanding flood risk and protecting communities. Today, this data is more abundant than ever, said NOAA Social, Behavioral, and Economic Sciences Data Engagement Coordinator Jonathon Mote, who presented on the importance of data integration to answering key questions about flooding. Yet, as he also emphasized, the sheer volume of data in different formats and from different sources poses challenges for access and usability.
“There’s been a data explosion, so there’s got to be some way to make sense of all the data and of how all the data connects to one another in the real world,” Mote said.
Key to this Workshop was the collective development of a knowledge graph. Knowledge graphs, which originate from the world of artificial intelligence and computer science, provide a way to represent, retrieve, and integrate data from disparate sources. In the world of disaster mitigation and risk management, a knowledge graph can map relationships between elements that influence flood risk, from people and infrastructure to weather patterns and government policies.
Natural Hazards Center Assistant Director Jennifer Tobin explained, “Creating knowledge graphs during this workshop encouraged participants to move outside of their disciplinary silos and begin working together to answer pressing flood-related questions from a variety of perspectives.”
Keynote speaker Lilit Yeghiazarian is the lead investigator of the National Science Foundation-funded Urban Flooding Open Knowledge Network (UKOKN), a first-of-its-kind “information infrastructure” that uses knowledge graphs to pull together complex data about the built and natural environments and better understand and forecast the complex impacts of urban flooding.
“We’re thinking of convergent problems that are highly complex and are of extraordinary societal importance—flooding is one of them,” said Yeghiazarian in her keynote address. "Flood impacts ripple cascade across complex geographies and social, environmental and technical systems. But, because data that has been collected in individual sectors is often siloed, we have very limited understanding of these cascades.”
In her remarks, Yeghiazarian emphasized that knowledge graphs can be used to address this gap, providing local government officials, emergency managers, residents, and others with actionable flood information.
Improving data integration and usability is foundational to help researchers, policymakers, practitioners, and residents of communities at risk to flooding establish a shared understanding. Actionable flood data can enhance early warning capabilities, inform flood management strategies, increase community resilience and ultimately save lives.
This is more important now than ever, said Natural Hazards Center Director Lori Peek, citing the deadly July 2025 Central Texas floods along the Guadalupe River.
“In this world that we’re living in now where flood losses are
skyrocketing, we are sentinels. We must use integrated data to ring the alarm bells,” Peek said. “The problems are becoming more complex, the stakes are becoming higher, and the only way we’re going to make progress on the problems we face is by coming together.”