Estimating aboveground biomass of tropical urban forests with UAV-borne hyperspectral and LiDAR data

Urban trees and forests can contribute to climate change mitigation by sequestering carbon in their living tissues, with aboveground biomass (AGB) playing a pivotal role. This study explores the capability of UAV-borne hyperspectral and LiDAR data for estimating AGB in tropical urban forests. Structural attributes of trees, such as diameter at breast height (DBH), total height, and wood density, were collected from over 5600 individuals, forming a comprehensive AGB dataset. Our methodology included two primary AGB estimation strategies: an area-based strategy that correlated AGB with hyperspectral and canopy height data across various grid sizes and an individual tree crown (ITC)-based method that integrated canopy height, spectral signatures of individual trees and crown area. The findings indicate that increasing the grid size from 10 m to 50 m improved the R2 from 0.24 ± 0.04 to 0.61 ± 0.13, mainly due to reduced border effects. Furthermore, integrating canopy height and hyperspectral data enhanced the R2 of AGB estimates from 0.61 ± 0.13 to 0.70 ± 0.09 for a 50 × 50 m grid. Crucially, wavelengths centered at the green peak and red-edge were identified as key bands in AGB retrieval. Integrating hyperspectral and LiDAR data did not significantly enhance results for individual trees, where AGB was closely linked to tree height and crown area. This study underscores the potential of utilizing integrated UAV-borne sensors for biomass assessment in urban forest settings.