Liu, Chenxi2024-09-232024-09-232024-09-232024-09-09https://hdl.handle.net/10012/21066Forest carbon storage estimation is critical for global climate change mitigation efforts, as forests play a vital role in the carbon cycle. This study investigates the accuracy of using Remotely Piloted Aircraft (RPA)-based LiDAR for estimating tree Diameter at Breast Height (DBH) and carbon storage through 3D tree modeling techniques. Two Quantitative Structure Models (QSM) — TreeQSM and AdQSM were used to virtually reconstruct trees and estimate tree and forest carbon stocks, with comparisons made to in-situ DBH measurements and traditional Allometric Scaling Models (ASMs), including self-developed allometric equation database and i-Tree Eco. Data were collected from two forest sites with different tree densities in Ontario, Canada, under leaf-on and leaf-off conditions. Result indicates that AdQSM generally outperformed TreeQSM in estimating DBH, particularly in less dense forests and during leaf-off season, where the correlation with in-situ DBH measurements improved. Despite variation in results from the two models, RPA-based LiDAR demonstrated potential as a scalable and non-destructive tool for forest carbon estimation, providing a foundation for future advancements in localized ASMs. The findings highlight the importance of accurate tree parameter extraction for forest carbon accounting, aligning with global efforts in climate change mitigation and sustainable forest management.enforest carbon storageQuantitative Structural Models3D tree modelingLiDARforest aboveground biomassAllometric Scaling ModelMaster Thesis