TUSCALOOSA, Ala. — As artificial intelligence reshapes our world, a University of Alabama researcher is tackling one of the technology’s most overlooked challenges: its massive thirst for water.
Dr. Jonathan Frame, an Alabama Water Institute and CIROH Faculty Fellow, has secured a $1.02 million National Science Foundation grant to investigate how expanding AI infrastructure will impact water resources—a critical issue he believes will define the next decade of water management.
“I’ve become more and more convinced that water for AI is going to be one of the biggest water resource issues of the next decade,” said Frame, assistant professor in UA’s Department of Geological Sciences. “We are excited to take on this challenging project.”
The three-year award, titled “AI-Guided Water Availability Tracking and Twin Systems for Infrastructure Resilience,” represents a pioneering effort to understand and manage the water demands of our increasingly digital world. The research addresses a pressing national need: enabling technological growth while ensuring the long-term availability and quality of water resources.
Physics-Informed AI Meets Water Science
The grant, funded through NSF’s Collaborations in Artificial Intelligence and Geosciences program, brings together expertise from multiple UA departments. Frame will lead a team that includes co-principal investigators Dr. Fei Hu from electrical and computer engineering, Dr. Yong Zhang from geological sciences and Dr. Jia Zhao from mathematics.
What sets this research apart is its integration of physics-informed artificial intelligence with hierarchical digital twin technologies. This novel approach promises to deliver both theoretical advances in AI algorithms tailored specifically to geoscience applications and practical guidance for environmental management strategies.
A Three-Pronged Scientific Approach
Frame’s project tackles the challenge through three main technical components that span from continental to local scales. First, his team will develop an AI-driven hydrologic model to analyze geospatial data across the contiguous United States, identifying regions with adequate water resources to support new AI infrastructure.
The second component involves creating digital twins for selected sites—sophisticated computer models that enable “what if?” scenario analyses to understand potential impacts from natural hazards and water availability fluctuations.
Finally, the research will employ fractional-calculus based modeling to assess environmental impacts associated with pollution from data centers, providing crucial insights into the long-term sustainability of AI infrastructure expansion.
The research methodology spans from broad regional analysis to site-specific hazard assessment.
From Theory to Practice
The research team’s interdisciplinary approach extends beyond academia through collaborations with national laboratories and water management agencies, ensuring that findings translate into practical solutions for real-world challenges.
Frame’s work addresses the complex reality that as data centers proliferate to meet growing AI demands, they create new pressures on water resources that existing planning frameworks weren’t designed to handle. His methodology will help identify suitable locations where water resources can reliably support AI infrastructure growth without environmental degradation or resource depletion.
During the proposal development process, Frame collaborated with experts across water security disciplines, including Mike Gremillion, director of AWI’s Global Water Security Center.
“Our conversation really helped to put the technical projects in context of large-scale water issues,” Frame noted, crediting Gremillion’s insights for helping frame the research within broader national security and resource management concerns.
Pioneering an Emerging Field
The timing of Frame’s research addresses a critical gap in infrastructure planning. While AI applications expand from smartphones to autonomous vehicles to smart cities, the water implications of this digital transformation have received little scientific attention.
Frame’s team will produce practical recommendations for AI investors, local agencies and site developers, providing policymakers and industry leaders with evidence-based information to manage resources effectively, mitigate conflicts and promote sustainable development practices.
The research promises to deliver novel methods for large-scale water resource analysis and new AI algorithms specifically designed for geoscience applications—tools that could reshape how we plan digital infrastructure nationwide.
Looking Ahead
The project, running from October 2025 through September 2028, will produce a comprehensive hydrologic modeling framework that could transform how the nation approaches technology infrastructure planning. The research combines expertise in hydrology, AI and applied mathematics to tackle one of the most complex resource management challenges of the digital age.
For Frame, this NSF award represents an opportunity to establish an entirely new research field at the intersection of artificial intelligence and water science. The project’s broader impacts extend far beyond academic publications—it aims to provide practical tools for sustainable development in an era of unprecedented technological growth. The research promises to position UA and AWI at the forefront of addressing a challenge that touches every aspect of modern life.
——-
References, graphic center image, data point origins – https://www.nature.com/articles/s41545-021-00101-w