Soil carbon effectively measured by new, efficient AI model

A new computer model is one of the first artificial intelligence tools to advance scientific discovery in agriculture and biogeochemistry and is 50 times more efficient than its predecessors, according to a study that offers a proof-of-principle for how AI might be used to shed light on obscure biological processes.

In a paper published July 24 in the journal Geoscientific Model Development, the researchers demonstrated the AI on processes behind the important issue of soil organic carbon, as the Earth's soils hold roughly three-quarters of the world's terrestrial carbon and more carbon than the atmosphere and all the world's plants combined.

Scientists have been exploring ways to use AI for research purposes, but most common AI tools, such as ChatGPT, mainly repurpose existing information. Researchers have also used AI to extract patterns from data. But the new model, called the Biogeochemistry-Informed Neural Network (BINN) goes a step further by predicting biological processes that are not yet well understood and suggesting factors that control 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."

Haodi Xu, a doctoral student in Luo's lab, is co-first author of the study. The research was a collaboration with the lab of Carla Gomes, the Ronald C. and Antonia V. Nielsen professor of computer science in Cornell Bowers Computing and Information Science, and professor in the Cornell SC Johnson College of Business.

Accurately understanding soil carbon processes could make a big difference in both predicting climate change and in developing practical solutions, and even small adjustments to these processes can have oversized downstream effects. For example, in 2015 French soil scientists proposed the "4 per 1,000" initiative, which claimed in theory that if humans adopted practices to increase global soil organic carbon in agricultural lands by 0.4% annually, it would not only improve soil health but also offset all the human-caused carbon emissions.

Soil scientists know the mechanisms by which soils acquire organic carbon- plants extract and sequester carbon from carbon dioxide to grow, and when those plants die, organic matter from stems, leaves and roots decompose into smaller and smaller bits to become part of the earth. But what is not well known are the speed of these processes and how many such processes are required to break down the litter.

"We use AI and data to tell us quantitatively how fast and how many of these kinds of processes are required," Xu said.

When compared to previous models, BINN computed 50 times faster. The accuracy of predictions of quantities of soil organic carbon was found to be very similar to the previous models. But previous models contained spatial biases, meaning that when making predictions across the contiguous U.S., it might favor the data from one area versus another. The researchers found less spatial bias with BINN.

The new model can be adapted to reveal other little-known agricultural and biogeochemical processes, such as those relating to soil respiration or carbon accumulation in forest systems, researchers say.

Co-first authors include Joshua Fan, a former doctoral student in the field of computer science in Gomes' lab who now works at Google, and Feng Tao, a former Schmidt AI for Science Postdoctoral Fellow at Cornell who is now an assistant professor at Nanyang Technological University, Singapore. Benjamin Z. Houlton, the Ronald P. Lynch Dean of the College of Agriculture and Life Sciences, and professor in the Department of Ecology and Evolutionary Biology and in the Cornell CALS Ashley School, is a co-author. Other co-authors from Cornell University include Lifen Jiang, Ying Sun, and Fengqi You.

The research was funded by the U.S. Department of Agriculture, the National Science Foundation, the Schmidt Sciences programs, the U.S. Department of Energy, CALS, the New York State Department of Environmental Conservation and the Air Force Office of Scientific Research.