By Blake Jackson
A newly developed artificial intelligence model is helping scientists better understand complex biological processes in agriculture and biogeochemistry while operating up to 50 times faster than earlier systems.
The breakthrough, detailed in a study published July 24 in Geoscientific Model Development, highlights how AI can move beyond analyzing existing data to support new scientific discoveries.
Researchers tested the model by studying soil organic carbon, a critical component of the global carbon cycle. Earth’s soils store nearly three-quarters of the planet’s terrestrial carbon, exceeding the amount found in the atmosphere and all plant life combined. Better understanding how carbon behaves in soil could improve climate predictions and guide sustainable agricultural practices.
Unlike many popular AI tools that primarily reorganize existing information, the new system—called the Biogeochemistry-Informed Neural Network (BINN)—is designed to predict biological processes that remain poorly understood while identifying the factors that influence them.
“BINN is very easy to use and can be democratized among the scientific community in various disciplines,” said Yiqi Luo, the Liberty Hyde Bailey Professor in the School of Integrative Plant Science, Soil and Crop Sciences Section, in the College of Agriculture and Life Sciences (CALS) and a senior author of the study.
“This is one of the first tools of this type that can promote scientific research with AI.”
The research was led by doctoral student Haodi Xu in collaboration with experts from Cornell University’s computer science and business programs.
Scientists already understand that plants absorb carbon dioxide and store carbon as they grow, while decomposing plant material eventually becomes part of the soil. However, many questions remain about how quickly these processes occur and how many stages are involved.
“We use AI and data to tell us quantitatively how fast and how many of these kinds of processes are required,” Xu said.
Compared with previous models, BINN produced similar prediction accuracy while operating 50 times faster and reducing geographic bias across the United States.
Researchers believe the model could also improve studies of soil respiration, forest carbon storage, and other agricultural and environmental processes.
Photo Credit: gettyimages-sasiistock
Categories: New York, Education, General