AI to Assess Soil Carbon Stocks
Researchers from Kazakhstan and Russia have signed an agreement to develop an AI-powered system for analyzing the condition of agricultural land and forecasting soil carbon stocks.
Agriculture is one of the largest sources of greenhouse gas (GHG) emissions, accounting for about 22% of global emissions, according to the United Nations. At the same time, cultivated soils absorb carbon from the atmosphere and store it as organic matter. Carbon is also a key component of humus, which determines soil fertility. Analyzing the carbon balance in soil is therefore important for forecasting crop yields and assessing the environmental impact of farming practices.
Joint Research by Russia and Kazakhstan
Researchers from Russia and Kazakhstan will use artificial intelligence to forecast changes in soil organic matter stocks while also assessing greenhouse gas emissions. The agreement on joint research was signed by the National Academy of Sciences of Kazakhstan and the Federal Research Center for Agroecology of the Russian Academy of Sciences.
The two sides have approved a development roadmap. The analysis will draw on satellite imagery, soil data and agrometeorological information. AI will help assess the carbon balance and the effectiveness of different soil management practices. Based on conditions in individual fields, researchers will be able to model greenhouse gas fluxes, forecast changes in soil composition and evaluate the performance of different agricultural technologies.
The collaboration began last year. On October 30, 2025, the 70-hectare Kaz Agro Carbon carbon research site became operational. It was established at the A.I. Barayev Kazakh Research and Production Center for Grain Farming in partnership with the Federal Research Center for Agroecology of the Russian Academy of Sciences. The site provides the facilities needed to conduct the full range of research.
Advancing Russian Technologies
Russia has been developing digital technologies to analyze soil conditions and their environmental impact for several years.
For example, researchers at Tomsk State University (TSU), working with IT company Syncretis, have developed a neural network that uses satellite imagery to assess the condition of agricultural land. Soil scientists studied fields in different climate zones and identified relationships between the reflectance of the land surface and its spectral signatures in satellite images. Using these data, the AI system identifies areas with differing soil characteristics, speeding up researchers’ work.
“Assessing soil conditions becomes extremely difficult when the task involves not small farms but large agricultural areas covering tens or even hundreds of thousands of hectares. Software for field monitoring will significantly improve the efficiency of agricultural resource management,” said Oleg Merzlyakov, head of the Department of Soil Science and Soil Ecology at TSU’s Institute of Biology.
Meanwhile, researchers at the Faculty of Soil Science at Lomonosov Moscow State University (MSU) have developed a dynamic carbon modeling system to inventory carbon stocks and greenhouse gases in croplands. The technology makes it possible to use experimental monitoring data from individual study sites at carbon research stations and assess whether proposed carbon sequestration technologies are justified.
Carbon Analytics for the World
The new platform being developed by Russia and Kazakhstan will enable high-quality monitoring and analysis of greenhouse gas emissions, making it possible to predict the risks of environmental damage and land degradation.
The scientific community will gain an additional AI tool for advancing soil science. This is particularly important given the vast areas of agricultural land in both countries, whose productivity must be maintained as the climate changes. The technology is also designed to help improve the quality of decision-making in land management. That, in turn, could give crop production a new boost and strengthen food security in the countries participating in the project.
In the future, the platform could also become an export product for countries seeking to protect the environment and meet international environmental requirements, including targets for reducing greenhouse gas emissions and developing projects that remove these gases from the atmosphere.
