AI Breakthrough: Forest Monitoring Revolutionized with Lidar-Level Accuracy from Satellite Images

Researchers from Beijing Forestry University, Manchester Metropolitan University, and Tsinghua University have developed a groundbreaking artificial intelligence model that revolutionizes forest monitoring by creating high-resolution canopy height maps using only standard RGB satellite imagery. Published in the Journal of Remote Sensing on October 20, 2025, this innovative approach combines large vision foundation models (LVFMs) with self-supervised learning to achieve near-lidar accuracy at a fraction of the cost. The model demonstrated remarkable precision with a mean absolute error of just 0.09 meters and R² of 0.78 when compared to airborne lidar measurements, outperforming traditional CNN and transformer-based methods while enabling over 90% accuracy in single-tree detection.

The research team tested their AI framework in Beijing's Fangshan District, where it successfully mapped fragmented plantations of Populus tomentosa, Pinus tabulaeformis, and Ginkgo biloba using one-meter-resolution Google Earth imagery. The model captured subtle variations in tree crown structure that existing global models often missed, supporting individual-tree segmentation and plantation-level biomass estimation with R² values exceeding 0.9 for key species. When applied to geographically distinct forests in Saihanba, the network maintained robust accuracy, confirming its cross-regional adaptability and potential for national-scale carbon accounting. This innovation bridges the gap between expensive lidar surveys and low-resolution optical methods, offering a scalable solution for long-term carbon sink monitoring and precision forestry management.

Dr. Xin Zhang, corresponding author at Manchester Metropolitan University, emphasized that their model demonstrates how large vision foundation models can fundamentally transform forestry monitoring by combining global image pretraining with local self-supervised enhancement. The AI-based mapping framework offers a powerful and affordable approach for tracking forest growth, optimizing plantation management, and verifying carbon credits under initiatives like China's Certified Emission Reduction program. Future research will extend this method to natural and mixed forests, integrate automated species classification, and support real-time carbon monitoring platforms. As the world advances toward net-zero goals, such intelligent, scalable mapping tools could play a central role in achieving sustainable forestry and climate-change mitigation, with the research findings accessible through the Journal of Remote Sensing and related resources like Chuanlink Innovations.