Background In arid and semi-arid regions, such as the Ulan Buh Desert, strong winds and sandstorms pose significant threats to the stability of ecosystems and agricultural productivity. Farmland shelterbelts are a proven strategy to mitigate wind erosion and enhance the microclimate, thereby protecting crops from adverse weather conditions. This study aims to provide a comprehensive evaluation of the windbreak effects of different structural configurations of Populusalba var. pyramidalis shelterbelts in the Ulan Buh Desert oasis, offering scientific insights for optimizing shelterbelt design.

Methods A combination of numerical simulation and field measurement were employed to assess the windbreak performance of three types of shelterbelts: Single-row, two-row, and four-row configurations. Through the integration of ground-based lidar point cloud data into voxel models, high-precision 3D shelterbelt models were successfully constructed. Wind field simulations were performed using Phoenics software, and the reliability and accuracy of the simulated results were validated against field-measured data. The characteristics of the wind field around shelterbelts with different structures were investigated, changes in wind speed along horizontal and vertical directions and the effective protection distance (H denotes average tree height, H = 30 m) were analyzed, and the protective effect of the shelterbelts was evaluated.

Results 1) The accuracy of wind field simulation results is high, and the model fitting degree R² is > 0.90. 2) The wind field around the shelterbelts was characterized by six distinct zones: Wind speed attenuation on the windward side, acceleration above the shelterbelt, acceleration on both sides, weak wind conditions behind the shelterbelt, vortex formation, and wind speed recovery downstream. The relative wind speed changes showed a pattern of initial decrease followed by an increase in both horizontal and vertical directions. 3) The horizontal direction demonstrated the minimum wind speed at 2 H on the leeward side, while the vertical direction showed the maximum attenuation rate at 0.5 H. The effective protection distance for each shelterbelt configuration ranged from 13 to 15 H, with the four-row shelterbelt achieving the longest average protection distance of 15.84 H. The windbreak distance of the four-row shelterbelt increased by 2.26 H compared with the single-row, but only increased by 0.66 H compared with the two-row. Considering the ecological benefits, economic benefits and construction costs, the protective effect of the second row forest belt is the best.

Conclusions The voxelization method based on point cloud data are proved to be highly effective in simulating the wind field around shelterbelts, providing a robust model for detailed analysis. By evaluating the windbreak effect of three farmland shelterbelts, it provides a scientific basis for the optimal allocation of farmland shelterbelts in desert oasis.