Kyoto, Japan -- Forests have been benefitting humanity since long before the health benefits of forest bathing were discovered. They are major carbon sinks that provide a wide range of ecosystem services, including timber and non-timber forest products, recreation, and climate regulation.
Accurately assessing forest biomass is essential for understanding carbon storage and supporting sustainable forest management, but forests are vast three-dimensional structures and therefore difficult to study. Until recently, even measuring the height of a single tree was a challenging task, let alone understanding the size of its canopy. Conventional ground-based tree surveys are labor-intensive and often difficult to conduct in remote or steep terrain, limiting their use in large-scale assessments. This has also restricted researchers' ability to develop accurate biomass estimation formulae.
However, new drone-based technologies such as LiDAR, or Light Detection and Ranging, are becoming increasingly accessible to researchers and have enabled more efficient measurement of tree structures and forest biomass. Emitting hundreds of thousands to millions of laser beams per second, LiDAR obtains three-dimensional information about the objects it scans.
A team of researchers at Kyoto University took advantage of LiDAR to conduct a large-scale forest survey of Japan. They measured the crown structure of 4,326 canopy trees representing 149 tree species across 23 forest census plots throughout the country: from sub-boreal forests in Hokkaido to subtropical ones in Okinawa. The team then integrated the drone-derived data with detailed ground-based measurements and developed species-specific equations to estimate tree biomass based on crown structural traits.
"To our knowledge, this represents the most comprehensive study to date estimating biomass for such a large number of tree species using drone-based data," says first author Kyaw Kyaw Htoo.
The analysis revealed that a model using only tree height and crown area could account for 72% of the variation in biomass. When incorporating functional types, for example conifer, deciduous broadleaf, and evergreen broadleaf, the explanatory power increased to 79%, and further improved to 83% when species-level information was included.
"Though drones cannot directly capture understory trees, our study found that canopy trees account for about 75% on average of total forest biomass across diverse forest types," says team leader Yusuke Onoda. "These results provide a foundation for estimating total forest biomass, including the understory."
This approach enables more accurate biomass estimation in species-rich natural forests by providing a repeatable, scalable, and data-driven tool for long-term forest monitoring and evaluation, and even has the potential to improve carbon credit accuracy and biodiversity monitoring.
"By making good use of these tools, we aim to enhance the efficiency of forest resource assessment, and hope to contribute not only to scientific progress but also to biodiversity conservation and to promoting sustainable forestry," says Htoo.
