UAV-Based Lidar and Multispectral Data for Carbon Stock Assessment in a Temperate Mixed Forest

—This study investigated the use of unmanned aerial vehicle (UAV)-based LiDAR and multispectral data for carbon stock assessment in a temperate mixed forest in Northern Italy. A DJI Matrice 300 RTK UAV equipped with a LiDAR sensor, a high-resolution RGB camera, and a multispectral camera were used to collect data on individual tree detection, tree height, and species composition. The LiDAR data was used to create a canopy height model and detect individual trees, while the multispectral data was used to classify tree species using a random forest classification algorithm. Allometric equations were then used to estimate the carbon stock of each tree based on its height and species. The results showed that the UAV-based method was effective in detecting individual trees (F-score=0.74) and estimating their height (R2=0.94). The species classification also performed well, with an overall accuracy of 84%. The distribution of carbon stock values showed differences depending on the tree height and species composition. English oak showed significantly higher carbon stock values compared to white hornbeam and black locust. This methodology can be used to establish a baseline for carbon stock estimation and monitoring over time, which is essential for generating carbon credits and evaluating carbon sequestration rates.