Abstract: Protective forests are the ecological barriers of oases in arid sand areas and can effectively prevent and control wind and sand hazards. The structural characteristics of individual trees, as the basic unit of protective forests, are the key factors affecting the protective benefits. With the typical leafless tree species of Ulan Buh Desert oasis, i.e., Populus alba var. pyramidalis, Populus nigra var. thevestina, and Populus popularis, as the research objects, and by using the ground-based LiDAR and through computational fluid dynamics (CFD), we fully explored the structural characteristics of individual trees and their surrounding aerodynamic characteristics on the basis of real 3D models. We further established the relationship between structural parameters of individual trees and wind field index. The results showed that combining AdQSM and MeshLab to build tree models had high accuracy. The wind field around the individual trees could be roughly divided into six regions, including the attenuation zone of the windward side of the plant, the acceleration zone at the top of the plant, the eddy zone, the calm zone, the transition zone, and the recovery zone of leeward side of the plant. The pressure field around individual trees showed a gradual change of high pressure on the windward side to low pressure on the leeward side. Horizontally, in the range of 20% to 50% reduction in relative wind speed, the effective protection distances were 0.21H-1.51H, 0.20H-0.91H, and 0.25H-1.64H (H was the corresponding tree height) for P. alba var. pyramidalis, P. nigra var. thevestina, and P. popularis, corresponding to effective protection areas of 18-294, 15-227, and 18-261 m², respectively. The maximum wind speed decay rate in the vertical direction was at 0.3H height for P. alba var. pyramidalis and P. popularis, and was reflected at 0.5H height for P. nigra var. thevestina. The correlation and stepwise regression analysis of the single tree structure parameters with the wind field indicators clearly indicated that optical porosity and volume porosity dominated the protection effect. Among the wind field factors, the best regression models related to the porous coefficient were screened for three factors, including diameter at breast height, tree surface area, and optical porosity. The regression variables screened for effective protection distance and effective protection area differed among the classes.

Key words: LIDAR, 3D model construction, wind field simulation, protection benefit