George Mason scientists harness AI for snow water equivalent forecasting in western United States

Ziheng Sun, Research assistant professor, Center for Spatial Information Science and Systems (CSISS), Geography and Geoinformation Science, College of Science; Mingrui Liu, assistant professor, Computer Science, College of Engineering and Computing (CEC); and Keren Zhou, assistant professor, Computer Science, College of Engineering and Computing, are studying the dynamics of snow water equivalent (SWE). 

SWE measures the amount of water available in snow. 

In this project, the researchers will use graph neural network-based models with physics-based constraints and partial differential equations to complete their work for this study. By doing this, they will create more accurate and reliable SWE forecasts by capturing the detailed processes of snow accumulation and melt. 

They will use the GeoWeaver workflow management platform to make advanced artificial intelligence tools workflows accessible, reproducible, and reusable by snow researchers and practitioners. 

The researchers will also carry out a series of hackathon-style workshops to provide students and snow researchers with hands-on experience in AI-powered and SWE forecasting. 

Overall, the scientists seek to democratize access to AI research workflows and knowledge tools for snow researchers, foster interdisciplinary collaboration, and support sustainable water resource management, thereby enhancing scientists’ understanding of snow water resources and contributing to the broader discourse on climate change and water sustainability.

Sun, Liu, and Zhou received $518,245 from the National Science Foundation for this project. Funding began in Oct. 2024 and will end in late Sept. 2027. The team will collaborate with Nicoleta Cristea from the University of Washington and Annie Burgess from Montana State University.