Optimizing UAV-LiDAR flight parameters for improved canopy penetration and structural metrics in agroforestry

Atalaya-Marin N.; Sanchez-Fuentes T.; Goñas M.; Tineo D.; Taboada-Mitma V.H.; Oblitas-Troyes J.F.; Fernandez-Jibaja J.A.; Torres-Herrera P.A.; Cabrera-Hoyos H.; Cruz-Luis J.; Gómez-Fernández D. · ISPRS Open Journal of Photogrammetry and Remote Sensing (2026) · DOI: 10.1016/j.ophoto.2026.100160

Resumen

Light Detection and Ranging (LiDAR) sensors mounted on unmanned aerial vehicles (UAVs) have become effective tools for estimating biomass and carbon in complex agroecosystems. However, a gap remains in the literature regarding the most influential flight parameters for acquiring high-quality LiDAR data. Hence, this study aims to evaluate the influence of UAV–LiDAR flight parameters, specifically flight speed, altitude, sensor angle, and trajectory pattern, on elevation accuracy, point cloud density, canopy penetration, individual tree segmentation, and aboveground biomass (AGB) and carbon estimation in a cocoa (Theobroma cacao L.) agroforestry system. A total of 41 UAV–LiDAR flights were conducted with a Matrice 350 RTK equipped with a Zenmuse L2 sensor. Field measurements from 58 georeferenced trees were used to validate LiDAR-derived height and Diameter at Breast Height (DBH); AGB was estimated using an allometric equation and converted to carbon using a factor of 0.5. Results showed that the two lowest flight speeds (3 and 6 m/s) produced higher-quality data and lower elevation errors, while altitudes between 40 and 60 m provided better canopy penetration and more consistent structural metrics. Similarly, dual grid paths with dual scan angles (−90° and −45°) produced denser point clouds and improved accuracy, achieving 93.1% tree-segmentation accuracy and an R2 of 0.72 for DBH estimation. Biomass estimates ranged from 29.8 to 185.9 t, with more stable values (42.9–50.8 t) obtained from low-altitude flights using multiple passes and scan angles. Overall, a flight speed of 3 m/s, altitudes between 40 and 60 m, and multiple trajectories with scan angles of −90° and −45° provided the highest UAV–LiDAR data quality for structural assessment in the cocoa agroforestry system. The Authors This is an open access article under the CC BY-NC-ND license. http://creativecommons.org/licenses/by-nc-nd/4.0/