AI Model Uses Himawari-8 Satellite Data for Hourly Carbon Absorption Monitoring Across East Asia

Approximately 30% of global carbon dioxide (CO2) emissions are absorbed by terrestrial vegetation through photosynthesis. Researchers from the Ulsan National Institute of Science and Technology (UNIST) have recently introduced an innovative AI analysis technique that can predict this carbon uptake with unprecedented temporal precision, down to hourly intervals. This breakthrough is anticipated to play a significant role in climate change mitigation and the development of carbon-neutral policies. The project, led by Professor Jungho Im from UNIST’s Department of Earth Environmental Urban Construction Engineering, involves an AI model that estimates daily gross primary production (GPP) at hourly intervals. GPP is a critical metric that indicates the amount of CO2 absorbed by plants during photosynthesis, making it essential for understanding ecosystem carbon sequestration. The team's findings are detailed in a paper published in Remote Sensing of Environment. The model utilizes high-frequency data from the Himawari-8 geostationary satellite, which captures images at 10-minute intervals. According to Sejeong Bae, the lead author of the study, this frequent data collection allows the model to accurately track diurnal variations in photosynthesis. "In contrast to polar-orbiting satellites, which typically observe a specific location only once or a few times a day, the Himawari-8 satellite provides continuous coverage, enabling us to capture the subtle but important changes in plant activity throughout the day," Bae said. The AI model also integrates various meteorological data, with a particular focus on Aerosol Optical Depth (AOD). AOD is a measure derived from satellite observations that indicates the concentration of particulate matter in the atmosphere. These particles can absorb or scatter solar radiation, influencing the amount and quality of light available for photosynthesis. To better understand how the AI model makes its predictions, the researchers used SHapley Additive exPlanations (SHAP), an advanced explainable AI technique. The results indicated that AOD has the most significant impact during the morning and evening hours, when the sun is lower on the horizon. At these times, more scattered light increases the sensitivity of photosynthetic activity to atmospheric aerosols, which helps explain why AOD is such a crucial variable. Professor Im emphasized the practical applications of the new model: "Our approach can map the spatial and temporal dynamics of carbon absorption over East Asia at a 2-kilometer resolution, covering a full 24-hour cycle. This level of detail is invaluable for ecosystem carbon flux analysis, vegetation monitoring, and environmental studies focused on light and carbon interactions." The potential implications of this research are substantial. By providing hourly data on photosynthesis, the model can help scientists and policymakers more accurately assess the impact of various environmental factors on carbon sequestration. This, in turn, can inform strategies for reducing atmospheric CO2 levels and achieving carbon neutrality, particularly in densely populated and industrially active regions like East Asia. Overall, the work spearheaded by Professor Im and his team represents a significant step forward in the application of AI to environmental science, offering a powerful tool for real-time monitoring and analysis of carbon absorption processes.