Accurate classification of forest stand is crucial for protection and management needs. However, forest stand classification remains a great challenge because of the high spectral and textural similarity of different tree species. Although existing studies have used multiple remote sensing data for forest identification, the effects of different spatial resolutions and combining multi-source remote sensing data for automatic complex forest stand identification using deep learning methods still require further exploration. Therefore, this study proposed an object-oriented convolutional neural network (OCNN) classification method, leveraging data from Sentinel-2, RapidEye, and LiDAR to explore classification accuracy of using OCNN to identify complex forest stands. The two red edge bands of Sentinel-2 were fused with RapidEye, and canopy height information provided by LiDAR point cloud was added. The results showed that increasing the red edge bands and canopy height information were effective in improving forest stand classification accuracy, and OCNN performed better in feature extraction than traditional object-oriented classification methods, including SVM, DTC, MLC, and KNN. The evaluation indicators show that ResNet_18 convolutional neural network model in the OCNN performed the best, with a forest stand classification accuracy of up to 85.68%.

Water resources in arid and semi-arid areas are limiting factors for ecosystem health and economic development. Therefore, an accurate and reasonable assessment of ecological water demand is crucial for efficient water resource utilization. In this study, we used vegetation coverage and groundwater depth to assess the state of vegetation growth in the Zhenglanqi, located at the southeastern edge of Otindag Sandy Land. Our results indicate the existence of a statistical power index function between vegetation coverage and groundwater depth scatter plots, where even minor changes in groundwater depth can have a significant impact on vegetation growth. In order to quantitatively assess the impact of subsidence on vegetation ecology, we propose a maximum allowable subsidence level under conditions that maintain normal ecological conditions, based on the initial subsidence depth and ecological guarantee rate. Our findings suggest that regions with shallower initial groundwater depths are more sensitive to changes in their environment than regions with deeper groundwater depths. The total groundwater consumption in the study area was 83 million cubic meters while maintaining an ecological guarantee rate of 80%; thus, while ensuring normal environmental conditions, human exploitation of shallow groundwater accounts for only 16 percent.