Urban tree carbon storage estimation using unmanned aerial vehicles remote sensing

Abstract

Urban green spaces play a crucial role in the carbon cycling process. The rapid remote sensing of carbon storage in trees within urban built environments remains a challenging endeavor due to the highly fragmented distribution of trees and complex land surface. The study aims to validate the feasibility of estimating individual urban tree carbon storage using UAV remote sensing and explore the driving factors of spatial variations in tree carbon storage. The diameter at breast height (DBH) of 841 Cinnamomum camphora (L.) Presl trees along six representative streets in Nanjing City was manually measured, and parameters including tree height (TH), crown area (CA), crown circumference (CC), and crown width (CW) were extracted from images of unmanned aerial vehicles (UAVs) using software such as DJI Terra, ArcGIS Pro, and Python. The results showed that the allometric growth relationship between urban tree DBH and TH was less stable than that of natural forests, which was related to the variation of the competition between individuals and building shading. In terms of determination coefficient (R²), UAV remote sensing could estimate urban tree DBH (R²=0.62) and carbon storage (R²=0.71) using multivariate estimation models in which all the input parameters was obtainable from UAV. Unexpectedly, the carbon storage of individual tree was positively correlated with competition intensity between canopies and building shading. This conclusion is useful to develop strategies to enhance carbon storage in urban green space, such as the appropriate increasing of tree planting density which can improve carbon storage of both individual tree and unit land area, and the more tree planting in densely built regions.