Geospatial scientist Hamdi Zurqani from the University of Arkansas at Monticello has pioneered a novel method to measure forest biomass using satellite data and artificial intelligence. Published in Ecological Informatics, this approach combines NASA’s spaceborne LiDAR with European satellite imagery to enhance the accuracy and efficiency of carbon cycle tracking—a critical tool in combating climate change.
Forests play a vital role in Earth’s carbon cycle, storing approximately 80% of terrestrial carbon. Traditional ground-based methods for estimating forest biomass are slow and labor-intensive, limiting their scalability. Zurqani’s research leverages open-access satellite data, processed via Google Earth Engine, to overcome these challenges.
The study integrated 3D canopy measurements from NASA’s GEDI LiDAR—mounted on the International Space Station—with optical imagery from the Copernicus Sentinel satellites. Four machine learning algorithms were tested, with gradient tree boosting delivering the highest accuracy in predicting biomass. Random forest also performed well, while support vector machines struggled, underscoring the importance of selecting the right AI tools for ecological analysis.
Zurqani emphasized that combining multi-source data, such as vegetation indices and topographic features, was key to achieving reliable results. This method enables large-scale biomass mapping, even in remote or inaccessible regions.
“Forests are the lungs of our planet,” said Zurqani. “This technology allows us to monitor their health and carbon storage potential with unprecedented precision, informing global climate policies.”
The study marks a significant advancement in environmental monitoring, though challenges like weather interference and limited LiDAR coverage persist. Future research may explore advanced AI models, such as neural networks, to further refine predictions. As climate change escalates, such innovations will be indispensable for sustainable forest management and climate action.
Add comment