Stand spatial structure of Coniferous forests based on multi-source LiDAR data

Coniferous forest spatial structure plays a critical role in forest management and ecological stability. However, traditional field survey methods for assessing stand spatial structure are labor-intensive, costly, and disruptive to ecosystems. To address these limitations, this study employs a multi-source LiDAR data fusion approach to explore intelligent methods for extracting stand spatial structure parameters.Using Larix principis-rupprechtii Mayr and Picea wilsonii as representative species, this study focuses on LiDAR data fusion as a core methodology to investigate intelligent approaches for extracting the spatial structure of forest stands in Shaanxi, China. The stand spatial structure parameters of the sample plot(including uniform angle index, neighborhood comparison and crowding degree) and the frequency were calculated and counted. The results showed that (1) in the coniferous forest stand with a total of 291 individual trees, the individual tree segmentation accuracy based on the integrated UAV-LiDAR and BLS data reached F = 0.96. (2) the Larix forest exhibited random distribution (R = 0.48), moderate size differentiation (U = 0.50), and average density (W = 0.75). The Picea crassifolia forest also showed random distribution (R = 0.48), moderate size differentiation (U = 0.47), and relatively high density (W = 0.94). The coniferous forest exhibited an unreasonable combination of tree distribution, with a frequency of 12%.The fused LiDAR data for parameter extraction and the calculation of forest stand spatial structure parameters enables faster and more effective analysis of spatial structure characteristics compared to traditional methods. Moreover, the multivariate distribution of these spatial parameters reveals internal structural features of the forest stand, allowing for the accurate identification of unreasonable structural combinations and providing a theoretical basis for optimizing and adjusting forest stand structure.Despite its promising results, this study is limited by the relatively small sample size and the specific forest types analyzed, which may constrain the generalizability of the findings.Future research should explore the integration of multi-temporal LiDAR datasets to assess dynamic changes in forest spatial structures and expand the methodology to diverse forest ecosystems.