AI has the potential to be more accurate than other models, but researchers still need more data.
The challenge of farm emissions
Agriculture presents a complex challenge in the fight against climate change. The United Nations recognizes its potential to store carbon, yet it’s a major source of greenhouse gasses globally.
Gasses emitted from farms, including carbon dioxide and nitrous oxide, contribute to climate change, according to the EPA. Despite the importance of measuring these emissions, accurately identifying their sources has proven difficult for scientists.
However, recent research from the University of Minnesota suggests that AI, with some human guidance, could provide a solution to this challenge.
Climate science and agricultural practices
Recent study by researchers at the University of Minnesota (UMN) explores the potential of AI to figure out how much pollution farms create.
This research was published in the science journal Nature Communications, on January 08, and addresses a long-time problem in climate science: precisely measuring how much greenhouse gas individual farms release.
Research indicates that combining process-based models, high-resolution data, and machine learning can address challenges in measuring carbon cycle dynamics.
Studies in the U.S. Corn Belt show that this integrated approach, called Knowledge-Guided Machine Learning (KGML), performs better than traditional methods, offering 86% more detailed insights into soil organic carbon changes.
This information helps scientists evaluate the effectiveness of different strategies against climate change and offers practical advice to farmers for minimizing their environmental impact.
This article was first published by Darija Gvozderac for Finbold. Read the original article here.
