Font Size

content body

Researchers collaborate on geospatial modeling that will help identify areas potentially vulnerable to PFAS exposure.

Researchers collaborate on geospatial modeling that will help identify areas potentially vulnerable to PFAS exposure.

Imagine having a time machine. 

You could travel back to a less stressful time. A time before concerns about chemicals in our soil and water — before farmers had to wonder whether an unseen contaminant could threaten an entire harvest or make its way into the food supply. 

We may not be able to turn back the clock, but Luyu Liu, an assistant professor in the Department of Geosciences, believes we can use technology to look ahead. 

Liu wants to leave his mark on the world. He wants to leave it a little better than he found it. 

His research is tackling one of today’s most persistent environmental challenges: per- and polyfluoroalkyl substances (PFAS), commonly known as “forever chemicals.” 

PFAS are a large group of human-made chemicals found in products and industrial applications, including firefighting foams and nonstick cookware. They do not readily break down and can move through soil and water, raising concerns about potential health risks, including cancer and fertility issues. 

Liu recently received a grant from the University of Maine PFAS Center of Excellence for his project, Physics-Informed AI for Hyperlocal PFAS Exposure and Agricultural Vulnerability Assessment. Working with researchers at the University of Maine, Liu will use artificial intelligence and environmental science to better understand where PFAS may pose a risk to agricultural land. 

PFAS can travel through water, meaning contamination may not remain where it originally entered the environment. It can move through rivers, streams and other waterways, potentially impacting agricultural soils and food sources. 

 

Liu recently received a grant from the University of Maine PFAS Center of Excellence for his project, Physics-Informed AI for Hyperlocal PFAS Exposure and Agricultural Vulnerability Assessment. Working with researchers at the University of Maine, Liu will use artificial intelligence and environmental science to better understand where PFAS may pose a risk to agricultural land. 

PFAS can travel through water, meaning contamination may not remain where it originally entered the environment. It can move through rivers, streams and other waterways, potentially impacting agricultural soils and food sources. 

For farmers, that creates a difficult question: How do you know if your land is vulnerable before you plant, harvest and send food to market? 

Testing every piece of agricultural land is expensive and impractical. Liu’s goal is to use existing measurements and environmental information to create a much more detailed picture of where PFAS may be present. 

Rather than identifying a broad region where PFAS may be present, Liu wants to develop high-resolution maps that can show differences in potential PFAS exposure from one area to another. 

Liu’s approach brings together the three pillars of his research: environmental science, transportation processes and GeoAI, or geospatial artificial intelligence. For this project, those disciplines work together to understand how PFAS moves through the environment and predict where they may accumulate. 

The research will account for physical processes such as water movement and gravity. Liu’s team will combine this information with existing PFAS measurements and geographic and environmental data. AI will then help fill in the gaps between the relatively small number of locations where researchers have actual measurements. 

“We enhance the very sparse observations on the ground into a continuous surface,” Liu said. 

In other words, rather than relying only on the places where PFAS has already been tested, Liu hopes to use AI to create a more complete picture of potential exposure across the landscape. 

Ultimately, the project will produce a PFAS Exposure Vulnerability Index and an AI-enabled dashboard designed to help identify areas where additional testing and monitoring may be most important. The technology could help farmers make informed decisions before investing time and resources into a crop. 

For Liu, however, the project is about more than developing a sophisticated computer model. 

“I want to do practical and very impactful research that can benefit our communities,” Liu said. 

That goal reflects Liu’s own multidisciplinary path. Trained as an environmental scientist, he later pursued GIS and now works at the intersection of environmental science and geospatial artificial intelligence. 

By bringing those fields together, Liu hopes technology can do more than explain problems after they happen. It can help communities anticipate them, understand them and respond. 

We may not be able to go back to a time before “forever chemicals,” but through research like Liu’s, we may be able to build a future where we are better prepared to protect the land, the people who depend on it and the food we put on our tables.