Canopy Structural Complexity (CSC) is a key structural attribute for evaluating forest ecological functions and health, with accurate canopy extraction serving as the prerequisite for its quantification and the basis for improving metric reliability and elucidating spatiotemporal canopy dynamics. However, existing canopy extraction methods generally rely on large-scale fully labeled point cloud datasets, which are often impractical for large-scale Unmanned Aerial Vehicle (UAV) LiDAR forest surveys. To address this limitation, we propose a weakly supervised canopy extraction strategy that relies on a limited number of labeled plots. By incorporating a pseudo-label generation strategy and a data-augmentation-based consistency self-supervised constraint, the method effectively improves segmentation performance in unlabeled scenarios. Furthermore, we introduce a multi-dimensional joint canopy entropy index that integrates both global and local canopy features, combining projection-based global canopy entropy with graph-based local canopy entropy to provide a more comprehensive CSC quantification. Experiments were conducted using three labeled plots as the training set and a held-out labeled plot for quantitative evaluation, with numerous unlabeled plots used for qualitative generalization analysis. Results showed that on the labeled test plots, the proposed weakly supervised strategy achieved 85.37 % mIoU and 92.64 % OA, surpassing the best-performing fully supervised method. Visual comparisons in real, unlabeled forest scenes demonstrate that the proposed method maintains clear boundaries between canopy and non-canopy regions, even in areas with significant topographic variation, fragmented canopy, or complex structures, while effectively suppressing false positives in ground and low-vegetation areas. The proposed multi-dimensional joint canopy entropy, when applied to the predictions of the weakly supervised canopy extraction strategy, successfully quantified the CSC using large-scale, unlabeled UAV forest LiDAR point clouds. It yielded results consistent with ecological priors and exhibited robustness under varying levels of point cloud sparsification. Based on normalized scores across three evaluation metrics, the proposed index achieved a score of 0.76, outperforming several commonly used CSC quantification indices.

Leaf nitrogen content (LNC) is a crucial indicator for assessing the nitrogen status of forest trees. The LNC retrieval can be achieved with the inversion of the PROSPECT-PRO model. However, the LNC retrieval from the commonly used leaf bidirectional reflectance factor (BRF) spectra remains challenging arising from the confounding effects of mesophyll structure, specular reflection, and other chemicals such as water. To address this issue, this study proposed an advanced BRF spectra-based approach, by alleviating the specular reflection effects and enhancing the leaf nitrogen absorption signals from Ginkgo trees and saplings, using 3 modified ratio indices (i.e., mPrior_800, mPrior_1131, and mPrior_1365) for the prior estimation of the N_struct structure parameter, combined with different inversion methods (STANDARD, sPROCOSINE, PROSDM, and PROCWT). The results demonstrated that the prior N_struct estimation strategy using modified ratio indices outperformed standard ratio indices or nonperforming prior N_struct estimation, especially for mPrior_1131 and mPrior_1365 yielding reliable performance for most constituents. With the use of the optimal approaches (i.e., PROCWT_S3 combined with mPrior_1131 or mPrior_1365), our results also revealed that the optimal estimation of LNC_area (normalized root mean square error [NRMSE] = 12.94% to 14.49%) and LNC_mass (NRMSE = 10.11% to 10.75%) can be further achieved, with the selected optimal wavebands concentrated in 5 common main domains of 1440 to 1539 nm, 1580 to 1639 nm, 1900 to 1999 nm, 2020 to 2099 nm, and 2120 to 2179 nm. These findings highlight marked potentials of the novel BRF spectra-based approach to improve the estimation of LNC and enhance the understanding of the impact of N_struct prior estimation on the LNC retrieval in leaves of Ginkgo trees and saplings.

Currently, China's forestry development still faces challenges such as uneven distribution of resources and overall low quality, which hinder the transformation of forestry development from focusing on quantity growth to quality improvement. In the face of major strategic needs such as national ecological security, timber security, food and oil security, and the “dual carbon” goals, it is urgently necessary to achieve accurate monitoring of tree phenotype traits, thereby selecting excellent germplasm resources, shortening breeding cycles, improving tree resilience, and wood quality. Traditional tree phenotype monitoring methods suffer from limited samples, low efficiency, poor accuracy, and sometimes even destructive processes, which constrain the efficiency and quality of tree breeding. The lack of efficient and accurate high-throughput phenotype information acquisition methods and analysis techniques has become one of the main bottlenecks hindering genetic analysis and fine breeding of trees. Modern unmanned aerial vehicle (UAV) remote sensing technology has the capability to intelligently, rapidly, and accurately capture dynamic changes in tree phenotype traits at multiple scales, which is of great significance for breaking through the bottleneck of tree phenotype monitoring mentioned above. Leveraging high-resolution passive and active remote sensing data obtained by UAV and intelligent analysis algorithms such as deep learning, machine learning, and data mining, it is possible to accurately extract multi-scale tree phenotype traits, providing a quantitative data guarantee for revealing the response relationships between “genes-phenotypes-environment”. This, in turn, facilitates further achievements in the selection of excellent germplasm resources, precision cultivation, identification and quantification of stresses, and resistance breeding. This paper first introduces the current application status of UAV remote sensing sensors in tree phenotype monitoring. Then, it focuses on the application progress of UAV remote sensing technology in extracting tree morphological structure traits, physiological functional traits, and biochemical component contents. Finally, it outlines the future development trends of UAV-based tree phenotype monitoring remote sensing technology from four aspects: multi-temporal and periodic dynamic monitoring of tree phenotype traits, integration of multi-source phenotype data obtained by UAVs, fusion of remote sensing data from different platforms of space-air-ground for collaborative monitoring and multi-omics analysis of tree phenotypes.

The diameter at breast height (DBH) of trees and stands is not only a widely used plant functional trait in ecology and biodiversity but also one of the most fundamental measurements in managing forests. However, systematically measuring the DBH of individual trees over large areas using conventional ground-based approaches is labour-intensive and costly. Here, we present an improved area-based approach to estimate plot-level tree DBH from airborne LiDAR data using the relationship between tree height and DBH, which is widely available for most forest types and many individual tree species. We first determined optimal functional forms for modelling height-DBH relationships using field-measured tree height and DBH. Then we estimated plot-level mean DBH by inverting the height-DBH relationships using the tree height predicted by LiDAR. Finally, we compared the predictive performance of our approach with a classical area-based method of DBH. The results showed that our approach significantly improved the prediction accuracy of tree DBH (R² ​= ​0.85–0.90, rRMSE ​= ​9.57%–11.26%) compared to the classical area-based approach (R² ​= ​0.80–0.83, rRMSE ​= ​11.98%–14.97%). Our study demonstrates the potential of using height-DBH relationships to improve the estimation of the plot-level DBH from airborne LiDAR data.