Tree height in tropical forest as measured by different ground, proximal, and remote sensing instruments, and impacts on above ground biomass estimates

doi:10.1016/j.jag.2019.101899

Abstract
Tree height is an important structural trait, critical in forest ecology and for above ground biomass estimate, and difficult to accurately measure in the field especially in dense forests, such as the tropical ones. The accuracy of height measurements depend on several factors including forest status, the experience of the observer, and the equipment used, with large subjectivity, heterogeneity and uncertainty in results, that can propagate when tree height is used in models. A comparison of Terrestrial Laser Scanning, Airborne Lidar Scanning, and stereo-photogrammetry (with imagery acquired by a RGB camera mounted on Unmanned Aerial Vehicle) approaches for estimating tree height was here performed, also with reference to ground methods. In fact, all those technique may increase the possibility of precise tree height measures, while reducing manual effort in comparison to more traditional ground techniques. The research was carried out in a dense tropical forest in Ghana; differences in measured heights as well as their impact on above ground biomass estimation were analyzed. All the different methods were characterized by pros and cons: the obtained results indicate that in dense forests, where sight occlusion problems occur, ground traditional techniques can lead to overestimation, while with the other mentioned techniques underestimation can occur, but in variable amount according to the considered instrument. The different height measures caused a remarkable variation in the estimated biomass of this tropical forest: more accurate height measurements are needed to reduce the uncertainty in biomass mapping efforts at any scale. Possibly, the simultaneous use of different methods can help in correctly estimate height uncertainty and reach a convergent and accurate result.

Keywords: Tree height; Forests; Biomass; UAV lidar; ALS; Photogrammetry; TLS