An alternative low-cost mobile LiDAR methodology for quantification of urban tree metrics

Urban trees provide a range of important ecosystem service benefits to society, including carbon storage and sequestration, flood mitigation and improving mental well-being. The delivery of these ecosystem services is largely dependent on the trees’ structural and functional traits. Manual surveys and allometric equations are commonly used to derive tree structural metrics (e.g., tree height, above-ground biomass); however, this approach is time-consuming and based on unsuitable allometric equations derived from rural trees, which will lead to uncertainty.

This study uses a low-cost mobile LiDAR sensor (MLS) system to quantify key structural metrics of urban trees. Using a Velodyne VLP-16 LiDAR scanner and a low-cost GPS unit, 197 transects, totalling 20 miles, were completed in park and street environments in Sheffield, UK. The data was processed using Simultaneous Localization and Mapping (SLAM) algorithms. Tree height and diameter at breast height (DBH) values were extracted utilising rlas (v1.6.2) and conicfit (v1.04) R packages. Quantitative volume metrics were extracted using quantitative structure models (QSM). A total of 80 urban trees and 32 species totalling 430 DBH, height and volume measurements were extracted from the MLS data.

MLS-derived results presented very strong agreement with manual field measurements (R² = 0.93, p < 0.001 and R² = 0.84, p < 0.001, for DBH and height, respectively). However, factors such as the slope of the terrain, occlusion and the distance from the tree contributed to varying levels of uncertainty in the results. Results using traditional allometric equations showed discrepancies with MLS-modelled above-ground biomass due to management controls on tree structure. Importantly, these findings point to the lack of transferability of rural allometry to urban trees and the importance of using techniques to repeatedly and accurately quantify the complete volumetric tree structure.