Forest canopy height reflects tree growth, biomass accumulation, and carbon storage potential. While global forest height maps now exist, most capture only a single moment in time, limiting their value for evaluating long-term forest dynamics. Field inventories provide accurate measurements but are expensive, labor-intensive, and difficult to scale across large or mountainous regions. In rapidly changing forest landscapes—especially those shaped by afforestation and plantation management—understanding growth trajectories over decades is essential for sustainable policy and climate planning. Based on these challenges, there is a need to conduct in-depth research into long-term, large-scale forest growth monitoring using consistent satellite observations.
Researchers from the Chinese Academy of Sciences, the University of Copenhagen, and collaborating institutions reported this work in the Journal of Remote Sensing, published (DOI: 10.34133/remotesensing.0810) in December 2025. The study addresses a key challenge in forest science: how to track forest growth continuously over long time periods and vast areas. By combining decades of Landsat satellite data with machine-learning models, the team reconstructed annual canopy height maps for southern China, offering a rare long-term perspective on how forests recover, mature, and respond to human management.
The analysis shows that forests in southern China grew substantially over the past three decades. Average canopy height increased from about 6.4 meters in 1986 to over 10.3 meters in 2019, representing a 61% rise. Areas dominated by taller trees expanded rapidly after 2000, reflecting large-scale afforestation and forest protection efforts. Plantation forests grew faster than secondary forests, with average height gains of about 0.20 meters per year, compared with 0.13 meters per year in secondary forests. However, secondary forests ultimately reached greater heights. The findings highlight that forest age and management practices, rather than climate alone, are the dominant drivers of long-term forest structure change.
To generate a continuous forest growth record, the researchers reconstructed annual canopy height maps at 30-meter resolution from 1986 to 2019. They trained a Random Forest machine-learning model using existing global forest height products and validated the results against national forest inventory data and airborne lidar measurements. Model accuracy remained stable across regions and decades, with average errors of about three meters.
The long-term dataset reveals distinct growth patterns. In the late 1980s and 1990s, landscapes were dominated by short trees. After 2000, canopy heights increased rapidly as planted forests matured. Plantation forests showed pronounced growth cycles linked to harvesting and replanting, while secondary forests displayed steadier, more stable growth. Statistical analyses further showed that forest age was the strongest factor controlling height changes, followed by precipitation and temperature, while soil properties limited maximum attainable height. Together, these results demonstrate that satellite spectral data can reliably capture vertical forest growth over time.
"This study shows that we can now observe how forests grow year by year, not just where they exist," one researcher noted. "By looking back more than 30 years, we can directly see how management decisions shape forest structure and carbon potential. This opens new possibilities for evaluating restoration success and guiding future forest policies."
The team analyzed the full Landsat satellite archive using Google Earth Engine, generating annual cloud-free composites. Multiple vegetation indices were combined with elevation data and processed through a Random Forest model trained on existing canopy height datasets. Predictions were validated using airborne lidar measurements and national forest inventory plots. Trend analysis, regression models, and driver attribution were then applied to quantify forest growth rates and identify the key environmental and management factors influencing canopy height changes.
This approach offers a scalable tool for monitoring forest growth, carbon accumulation, and management outcomes worldwide. With similar satellite archives available globally, the method can be applied to other regions undergoing reforestation or plantation expansion. As higher-resolution satellite and lidar data become more accessible, long-term forest monitoring from space could directly inform carbon accounting, biodiversity conservation, and climate strategies. Ultimately, the ability to watch forests grow from space may transform how societies manage and value forest ecosystems.
###
References
DOI
10.34133/remotesensing.0810
Original Source URL
https://spj.science.org/doi/10.34133/remotesensing.0810
Funding information
This work was supported by the National Key Research and Development Program of China for Young Scientists (2023YFF1305700), the National Natural Science Foundation of China (42371129), the National Natural Science Fund for Excellent Young Scientists (Overseas), the National Key Research and Development Program of China (2022YFF1300700), the Science and Technology Innovation Program of Hunan Province (2024RC1067), the International Partnership Program of Chinese Academy of Sciences (CAS) (092GJHZ2022029GC), and the CAS Interdisciplinary Team (JCTD-2021-16). M.B. acknowledges support from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement no. 947757 TOFDRY).
About Journal of Remote Sensing
The Journal of Remote Sensing, an online-only Open Access journal published in association with AIR-CAS, promotes the theory, science, and technology of remote sensing, as well as interdisciplinary research within earth and information science.