By Andi Anderson
A team of researchers from Michigan State University has introduced a method to better predict nitrous oxide emissions from U.S. croplands.
This new machine learning–based system offers far greater accuracy than previous approaches and could improve national greenhouse gas accounting and support more effective emission‑reduction strategies in agriculture. The study was published in the Proceedings of the U.S. National Academy of Sciences.
Nitrous oxide is a powerful greenhouse gas released mainly from nitrogen fertilizer use on farmland. Predicting its emissions has been difficult because many factors—including weather, soil conditions and crop practices—interact to influence how soil microbes produce the gas.
This new research aims to overcome those challenges. The project was led by former MSU graduate student Prateek Sharma and Hannah Distinguished Professor Bruno Basso from MSU’s Department of Earth and Environmental Sciences and the W.K. Kellogg Biological Station (KBS).
The team created a hybrid modeling system combining machine learning and ecosystem models to estimate daily nitrous oxide emissions with high precision.
The research effort was co‑led by G. Philip Robertson, University Distinguished Professor at KBS and in the Department of Plant, Soil and Microbial Sciences.
Professor Michael Murillo of the Department of Computational Mathematics, Science and Engineering also contributed to the project.
The model was trained using more than 12,000 nitrous oxide measurements collected across 17 sites in the U.S. Midwest and Great Plains. These measurements represented six cropping systems and 35 management practices, making the dataset one of the most extensive of its kind. Traditional models often achieve around 20% accuracy, but the new system reached more than 80% accuracy.
As Basso explained, “One of the limiting factors of current predictive models is that they rely on outdated national greenhouse gas emission inventories and often need to be calibrated to a specific site. With this effort, we’ve moved past these limitations to provide management-specific predictions for crucial combinations of cropping systems, soils, management practices and weather conditions. We're hopeful this approach can lead to field-specific emission mitigation strategies, as well as much-needed updates to estimates of greenhouse gas emissions from agriculture.”
Additional team members included Aditya Manuraj, Neville Millar, Tommaso Tadiello, Mukta Sharma and Mathieu Delandmeter from the Department of Earth and Environmental Sciences.
The project received support from several organizations, including the Great Lakes Bioenergy Research Center, the U.S. Department of Energy Office of Science, the National Science Foundation Long-Term Ecological Research Program at KBS, the USDA National Institute of Food and Agriculture, the USDA Long-term Agroecosystem Research Program at KBS, the CERCA-Foundation for Food and Agriculture Research Project, Climate Trace, the Soil Inventory Project and MSU AgBioResearch.
Photo Credit: michigan-state-university-msu
Categories: Michigan, Education