Research & Discovery

May 22, 2024 | Christine Buckley - College of Liberal Arts and Sciences

Geoscientist Among First Projects Approved by National Artificial Intelligence Research Resource (NAIRR) Pilot 

Lijing Wang, who joins UConn in August, will develop AI models for mountain water flow that aid in climate change predictions

Mountains in the distance, wetlands in the foreground.

The East River Watershed during an October 2023 research trip (Photo courtesy of Lijing Wang)

Lijing Wang, assistant professor of Earth Sciences in the College of Liberal Arts and Sciences, is among the first scientists in the U.S. to earn support from the National Artificial Intelligence Research Resource (NAIRR) Pilot, a nationwide infrastructure that connects U.S. researchers to the computational data, software, models, and training they need to conduct paradigm-shifting AI research. 

The U.S. National Science Foundation (NSF) and the Department of Energy (DOE) announced last week the first 35 projects awarded computational time through the project, marking what it calls a significant milestone in fostering responsible AI research and democratizing access to AI tools across the country. 

The NAIRR Pilot will support fundamental, translational and use-inspired AI-related research with emphasis on societal challenges. Initial priority topics include safe, secure and trustworthy AI; human health; and environment and infrastructure. 

Wang, who will join UConn as assistant professor of Earth Sciences in the fall, received 10,500 node hours at the DOE Argonne National Laboratory AI Testbed. A node hour is the cumulative amount of time that computing resources equivalent to one individual node, or a single computer within a larger network or cluster, have been active or utilized for computation. 

Her project studies water flow in mountainous areas where the lack of data about snow melt and water movement makes predictions about the area’s future water flow difficult to compute or inaccurate.  

“Mountainous watersheds provide significant water resources,” says Wang. “Conducting intensive monitoring is key to understanding water availability, but it’s not feasible in every catchment. Together with monitoring, an AI tool could help us evaluate these water variations more efficiently in the face of climate change.” 

The work will simulate water movement across multiple mountain slopes under different conditions, and the results will form a dataset for an AI model to predict snow melt, water flow, and groundwater levels. Her results will lead to more rapid water forecasting, which will improve water management and climate change studies. 

Of the 35 projects, 27 will be supported through the NSF-funded advanced computing systems, and eight projects will utilize DOE-supported systems.