Tree species information is an important factor in forest resource surveys, and light detection and ranging (LiDAR), as a new technical tool for forest resource surveys, can quickly obtain the 3D structural information of trees. In particular, the rapid and accurate classification and identification of tree species information from individual tree point clouds using deep learning methods is a new development direction for LiDAR technology in forest applications. In this study, mobile laser scanning (MLS) data collected in the field are first pre-processed to extract individual tree point clouds. Two downsampling methods, non-uniform grid and farthest point sampling, are combined to process the point cloud data, and the obtained sample data are more conducive to the deep learning model for extracting classification features. Finally, four different types of point cloud deep learning models, including pointwise multi-layer perceptron (MLP) (PointNet, PointNet++, PointMLP), convolution-based (PointConv), graph-based (DGCNN), and attention-based (PCT) models, are used to classify and identify the individual tree point clouds of eight tree species. The results show that the classification accuracy of all models (except for PointNet) exceeded 0.90, where the PointConv model achieved the highest classification accuracy for tree species classification. The streamlined PointMLP model can still achieve high classification accuracy, while the PCT model did not achieve good accuracy in the tree species classification experiment, likely due to the small sample size. We compare the training process and final classification accuracy of the different types of point cloud deep learning models in tree species classification experiments, further demonstrating the advantages of deep learning techniques in tree species recognition and providing experimental reference for related research and technological development.

To investigate forest resources, it is necessary to identify the tree species. However, it is a challenge to identify tree species using 3D point clouds of trees collected by light detection and ranging (LiDAR). PointNet++, a point cloud deep learning network, can effectively classify 3D objects. It is important to establish high-quality individual tree point cloud datasets when applying PointNet++ to identifying tree species. However, there are different data processing methods to produce sample datasets, and the processes are tedious. In this study, we suggest how to select the appropriate method by designing comparative experiments. We used the backpack laser scanning (BLS) system to collect point cloud data for a total of eight tree species in three regions. We explored the effect of tree height on the classification accuracy of tree species by using different point cloud normalization methods and analyzed the effect of leaf point clouds on classification accuracy by separating the leaves and wood of individual tree point clouds. Five downsampling methods were used: farthest point sampling (FPS), K-means, random, grid average sampling, and nonuniform grid sampling (NGS). Data with different sampling points were designed for the experiments. The results show that the tree height feature is unimportant when using point cloud deep learning methods for tree species classification. For data collected in a single season, the leaf point cloud has little effect on the classification accuracy. The two suitable point cloud downsampling methods we screened were FPS and NGS, and the deep learning network could provide the most accurate tree species classification when the number of individual tree point clouds was in the range of 2048–5120. Our study further illustrates that point-based end-to-end deep learning methods can be used to classify tree species and identify individual tree point clouds. Combined with the low-cost and high-efficiency BLS system, it can effectively improve the efficiency of forest resource surveys.

Introduction: The unmanned aerial vehicle -based light detection and ranging (UAV-LiDAR) can quickly acquire the three-dimensional information of large areas of vegetation, and has been widely used in tree species classification.Methods: UAV-LiDAR point clouds of Populus alba, Populus simonii, Pinus sylvestris, and Pinus tabuliformis from 12 sample plots, 2,622 tree in total, were obtained in North China, training and testing sets were constructed through data pre-processing, individual tree segmentation, feature extraction, Non-uniform Grid and Farther Point Sampling (NGFPS), and then four tree species were classified efficiently by two machine learning algorithms and two deep learning algorithms.Results: Results showed that PointMLP achieved the best accuracy for identification of the tree species (overall accuracy = 96.94%), followed by RF (overall accuracy = 95.62%), SVM (overall accuracy = 94.89%) and PointNet++(overall accuracy = 85.65%). In addition, the most suitable number of point cloud sampling of single tree is between 1,024 and 2048 when using the NGFPS method in the two deep learning models. Furthermore, feature value of elev_percentile_99th has an important influence on tree species classification and tree species with similar crown structures may lead to a higher misidentification rate.Discussion: The study underscores the efficiency of PointMLP as a robust and streamlined solution, which offers a novel technological support for tree species classification in forestry resource management.

Accurate tree species information is a prerequisite for forest resource management. Combining light detection and ranging (LiDAR) and image data is one main method of tree species classification. Traditional machinelearningmethods rely on expert knowledge to calculatea large number of feature parameters.Deep learning technology can directly use the original image and pointclouddata to classify tree species. However, data with different patterns require the use of different types of deeplearningmethods. In this study, a multimodal deeplearningframework (TSCMDL) that fuses 2D and 3D features was constructed and then used to combine data from multiple sources for tree species classification. This framework uses an improved version of the PointMLP model as its backbone network and uses ResNet50 and PointMLP networks to extract the image features and pointcloudfeatures, respectively. The proposed framework was tested using UAV LiDAR data and RGB orthophotos. The results showed that the accuracy of the tree species classification using the TSCMDL framework was 98.52%, which was 4.02% higher than that based on pointcloudfeatures only. In addition, when the same hyperparameters were used for training the model, the efficiency of the model training was not significantly lower than for models based on pointcloudfeatures only. The proposed multimodal deeplearningframework extracts features directly from the original data and integrates them effectively, thus avoiding manual feature screening and achieving more accurate classification. The feature extraction network used in the TSCMDL framework can be replaced by other suitable frameworks and has strong application potential.

Forest carbon flux is critical to climate change, and the accurate modeling of forest carbon flux is an extremely challenging task. The remote sensing model (the MODIS MOD_17 gross primary productivity (GPP) model (MOD_17)) has strong practicability and is widely used around the world. The ecological process (the Biome-BioGeochemical Cycles Multilayer Soil Module model (Biome-BGCMuSo)) model can describe most of the vegetation’s environmental and physiological processes on fine time scales. Nevertheless, complex parameters and calibrations pose challenges to the application and development of models. In this study, we optimized all the input parameters of the MOD_17 model for the calibration of the Biome-BGCMuSo model to obtain GPP with continuous spatiality. To determine the contribution of input parameters to the GPP of different forest types, an Extended Fourier Amplitude Sensitivity Test (EFAST) was performed on the Biome-BGCMuSo model firstly. Then, we selected the sample points of each forest type and its different ecological gradients (30 for each type), using the GPP simulation value of the optimized MOD_17 model corresponding to the time and space scale to calibrate the Biome-BGCMuSo model, to drive the calibrated Biome-BGCMuSo, and we simulated the different forest types’ net primary productivity (NPP). According to dendrochronological measurements, the NPP simulation results were verified on the whole regional scale. The results showed that the GPP values of different forest types were highly sensitive to C:N_leaf (C:N of leaf), SLA1 (canopy average specific leaf area in phenological phase 1), and FLNR (fraction of leaf N in Rubisco). The coefficient of determination (R²) between the simulated forest NPP and the measured NPP was 0.64, and the root-mean-square (RMSE) was 26.55 g/C/m²/year. Our study aims to reduce uncertainty in forest carbon fluxes simulated by the Biome-BGCMuSo model, providing feedback for understanding forest ecosystem carbon cycling, vegetation productivity, and climate change.

Tree species is a critical factor in the practice of forest resource field sample surveys. Light detection and ranging (LiDAR) can obtain three-dimensional structural information about forests and trees and is increasingly being used in forest resource surveys. We used three pointwise multi-layer perceptron (MLP)-based deep learning methods (PointNet, PointNet++, and PointMLP) to identify individual tree point clouds of seven different tree species to explore the effectiveness of point cloud deep learning in classifying individual tree point clouds. Experiment results were extremely exciting. Higher classification accuracy can be attained in trials utilizing 2048 points. The tree classification accuracies of PointMLP and PointNet++ on the test set were 0.9474 and 0.9483, respectively, in classification experiments with a balanced sample size. PointMLP, the current state-of-the-art pointwise MLP-based model, is faster to train and performs better.