TUM researchers create first global weekly indicator for plant health using hybrid AI and satellite data

TUM researchers create first global weekly indicator for plant health using hybrid AI and satellite data

(IN BRIEF) Researchers at the Technical University of Munich (TUM) have developed the first global, weekly indicator for plant health using a hybrid intelligence approach that combines physical models and artificial intelligence. The system relies on satellite data from Sentinel-3 to estimate canopy chlorophyll content (CCC), a key marker of plant metabolism and growth. Unlike conventional methods limited by surface reflectance data or cloud cover, the TUM approach delivers reliable information across landscapes ranging from farmland to forests. This indicator allows scientists to monitor dynamic changes in plant vitality, biomass production, and climate impacts, with applications for agriculture, climate adaptation, and global food security.

(PRESS RELEASE) MUNICH, 20-Aug-2025 — /EuropaWire/ — The Technical University of Munich (TUM) has unveiled a new global indicator for monitoring plant health that can provide reliable updates on a weekly basis. The method, developed by researchers at TUM’s School of Life Sciences, uses a hybrid intelligence approach combining physical models and artificial intelligence to generate worldwide data based on satellite imagery. This breakthrough offers an essential tool for science, agriculture, and climate planning.

The calculated value for plant health (Canopy Chlorophyll Content, short CCC) can be displayed globally on maps such as this one. The map shown is based on data from August 2024.

Until now, a consistent global dataset for plant health has been lacking, despite its importance for agriculture and climate research. The newly developed method, validated across a variety of landscapes, can assess vegetation health in agricultural fields, forests, and other land-use types. Even under challenging conditions such as partial cloud cover, the system is capable of producing accurate results by drawing on satellite images taken from the upper atmosphere.

The two-step process first applies a physical model to establish the link between chlorophyll levels and canopy reflectance. These results are then used to train AI models, enabling them to directly estimate canopy chlorophyll content (CCC) from satellite reflectance data. “By using this hybrid intelligence approach, our method overcomes the limitations of conventional methods that rely solely on reflectance data from the earth’s surface,” explained Dong Li, first author of the study.

Chlorophyll, a key molecule in photosynthesis, is closely tied to plant metabolism and growth. As such, its measurement serves as a reliable indicator of plant vitality, yield potential, and biomass production. With the new system relying on Sentinel-3 OLCI satellite data updated almost weekly, researchers can now observe how plants respond dynamically to changing conditions, including the effects of climate variability.

“Being able to track canopy chlorophyll content at a global scale gives us unprecedented insight into crop vitality and earth surface processes,” said Prof. Kang Yu, Chair of Precision Agriculture at TUM. “This data can help improve models for agriculture and climate adaptation, providing a solid scientific basis for decision-making worldwide.”

The team’s findings have been published in Remote Sensing of Environment. By delivering a global and regularly updated measure of plant health, TUM researchers are providing science and policy with a powerful new resource for understanding and adapting to agricultural and environmental challenges.

Publications

Dong Li, Holly Croft, Gregory Duveiller, Adam P. Schreiner-McGraw, Anirudh Belwalkar, Tao Cheng, Yan Zhu, Weixing Cao, Kang Yu, Global retrieval of canopy chlorophyll content from Sentinel-3 OLCI TOA data using a two-step upscaling method integrating physical and machine learning models, Remote Sensing of Environment, Volume 328, 2025, 114845, ISSN 0034-4257, https://doi.org/10.1016/j.rse.2025.114845.

Further information and links

Prof. Kang Yu is professor for the Chair of Precision Agriculture. His chair is affiliated with the TUM School of Life Sciences and the World Agricultural Systems Center – Hans Eisenmann-Forum for Agricultural Sciences.

Media Contacts:

Corporate Communications Center
Linda Schinnenburg
presse@tum.de

Contacts to this article:

Prof. Dr. Kang Yu
Technical University of Munich
Precision Agriculture Lab
+49 8161 71 5001
kang.yu@tum.de
https://www.lse.ls.tum.de/pag/home/

SOURCE: Technical University of Munich