Lehigh researchers say AI can help water systems plan
Lehigh University researchers are developing AI models to help water system managers better control floods, administer drought restrictions, and maintain hydropower operations amidst climate change. The federally funded project initially focuses on the Delaware River Basin but has nationwide applications. Data centers are noted as a significant and growing stressor on water resources, making such research more critical.
Lehigh University researchers are embarking on a federally funded project, dubbed RIVER-AI, to develop artificial intelligence-powered models to assist water system managers in navigating the challenges posed by drought and extreme storms. The initiative aims to help operators control floods, implement drought restrictions, and maintain hydropower operations by incorporating complex water rights and guidelines into AI model predictions. Dr. Hongyi Li, an associate professor of civil and environmental engineering, emphasized that the model's purpose is to enhance human decision-making, not replace it.
The project, which officially begins on September 1 and is funded by the U.S. Department of Energy, will initially focus on data from the upper Delaware River Basin, which supplies drinking water to over 14 million people across New York and Pennsylvania. Researchers anticipate nationwide applications for their work, planning collaborations with federal labs and other academic institutions. The AI approach will enable faster analysis of water systems under stress and allow for the modeling of scenarios that are too risky to test physically.
Brian Gramlich Rahm, director of the New York State Water Resources Institute at Cornell University, highlighted the increasing complexity of managing water resources, particularly with the rising demand from data center development. He noted the difficulty in obtaining reliable data on water usage by data centers and drew parallels to the scientific efforts during the fracking boom. While acknowledging the irony of using AI to address water strain when AI's power demands contribute to the problem, Gramlich Rahm stressed the urgent need for a societal discussion on resource priorities.