The segmentation of plant disease images enables researchers to quantify the proportion of disease spots on leaves, known as disease severity. Current deep learning methods predominantly focus on single diseases, simple lesions, or laboratory-controlled environments. In this study, we established and publicly released image datasets of field scenarios for three diseases: soybean bacterial blight (SBB), wheat stripe rust (WSR), and cedar apple rust (CAR). We developed Plant Disease Segmentation Networks (PDSNets) based on LinkNet with ResNet-18 as the encoder, including three versions: ×1.0, ×0.75, and ×0.5. The ×1.0 version incorporates a 4 × 4 embedding layer to enhance prediction speed, while versions ×0.75 and ×0.5 are lightweight variants with reduced channel numbers within the same architecture. Their parameter counts are 11.53 M, 6.50 M, and 2.90 M, respectively. PDSNetx0.5 achieved an overall F1 score of 91.96%, an Intersection over Union (IoU) of 85.85% for segmentation, and a coefficient of determination (R²) of 0.908 for severity estimation. On a local central processing unit (CPU), PDSNetx0.5 demonstrated a prediction speed of 34.18 images (640 × 640 pixels) per second, which is 2.66 times faster than LinkNet. Our work provides an efficient and automated approach for assessing plant disease severity in field scenarios.

As a major coal-producing area, the Shanxi section of the Yellow River Basin has been significantly affected by coal mining activities in the local ecological environment. Therefore, an in-depth study of the ecological evolution in this region holds great scientific significance and practical value. In this study, the Shanxi section of the Yellow River Basin, including its planned coal mining area, was selected as the research subject. An improved remotely sensed ecological index model (NRSEI) integrating the remotely sensed ecological index (RSEI) and net primary productivity (NPP) of vegetation was constructed utilizing the Google Earth Engine platform. The NRSEI time series data from 2003 to 2022 were calculated, and the Sen + Mann–Kendall analysis method was employed to comprehensively assess the ecological environment quality and its evolutionary trends in the study area. The findings in this paper indicate the following data: (1) The contribution of the first principal component of the NRSEI model is more than 70%, and the average correlation coefficient is higher than 0.79. The model effectively integrates the information of multiple ecological indicators and enhances the applicability of regional ecological environment evaluation. (2) Between 2003 and 2022, the ecological environment quality in the Shanxi section of the Yellow River Basin showed an overall upward trend, with the average NRSEI value experiencing phases of fluctuation, increase, decline, and stabilization. The NRSEI values in non-coal mining areas consistently remained higher than those in coal mining areas. (3) Over 60% of the areas have improved ecological conditions, especially in coal mining areas. (4) The impact of coal mining on the ecological environment is significant within a 6 km radius, while the effects gradually diminish in the 6 to 10 km range. This study not only offers a reliable methodology for evaluating ecological environment quality on a large scale and over a long time series but also holds significant guiding value for the ecological restoration and sustainable development of the Shanxi section of the Yellow River Basin and its coal mining area.

The geological environment damage caused by coal mining subsidence has become an important factor affecting the sustainable development of mining areas. Reconstruction of the Coal Mining Subsidence Field (CMSF) is the key to preventing geological disasters, and the needs of CMSF reconstruction cannot be met by solely relying on a single remote sensing technology. The combination of Unmanned Aerial Vehicle (UAV) and Synthetic Aperture Radar (SAR) has complementary advantages; however, the data fusion strategy by refining the SAR deformation field through UAV still needs to be updated constantly. This paper proposed a Prior Weighting (PW) method based on Satellite Aerial (SA) heterogeneous remote sensing. The method can be used to fuse SAR and UAV Light Detection and Ranging (LiDAR) data for ground subsidence parameter inversion. Firstly, the subsidence boundary of Differential Interferometric SAR (DInSAR) combined with the large gradient subsidence of Pixel Offset Tracking (POT) was developed to initialize the SAR preliminary CMSF. Secondly, the SAR preliminary CMSF was refined by UAV LiDAR data; the weights of SAR and UAV LiDAR data are 0.4 and 0.6 iteratively. After the data fusion, the subsidence field was reconstructed. The results showed that the overall CMSF accuracy improved from ±144 mm to ±51 mm. The relative errors of the surface subsidence factor and main influence angle tangent calculated by the physical model and in situ measured data are 1.3% and 1.7%. It shows that the proposed SAR/UAV fusion method has significant advantages in the reconstruction of CMSF, and the PW method contributes to the prevention and control of mining subsidence.

The Middle Route of the South-to-North Water Diversion Project is a critical infrastructure that ensures optimal water resource distribution across river basins and safeguards the livelihood of people in China. This study investigated its effects on the land surface temperature (LST) and fractional vegetation coverage (FVC) in the Danjiang River Basin. Moreover, it examined the spatial and temporal patterns of this project, providing a scientific basis for the safe supply of water and ecological preservation. We used the improved interpolation of mean anomaly (IMA) method based on the digital elevation model (DEM) to reconstruct LST while FVC was estimated using the image element dichotomous model. Our findings indicated a general increase in the average LST in the Danjiang River Basin post-project implementation. During both wet and dry seasons, the cooling effect was primarily observed in the south-central region during the daytime, with extreme values of 6.1 °C and 5.9 °C. Conversely, during the nighttime, the cooling effect was more prevalent in the northern region, with extreme values of 3.0 °C and 2.3 °C. In contrast, the warming effect during both seasons was predominantly located in the northern region during the daytime, with extreme values of 5.3 °C and 5.5 °C. At night, the warming effect was chiefly observed in the south-central region, with extreme values of 5.8 °C and 5.9 °C. FVC displayed a seasonal trend, with higher values in the wet season and overall improvement over time. Statistical analysis revealed a negative correlation between vegetation change and daytime temperature variations in both periods (r = −0.184, r = −0.195). Furthermore, a significant positive correlation existed between vegetation change and nighttime temperature changes (r = 0.315, r = 0.328). Overall, the project contributed to regulating LST, fostering FVC development, and enhancing ecological stability in the Danjiang River Basin.

Investigating the spatiotemporal variations in the Normalized Difference Vegetation Index (NDVI) in the Shanxi section of the Yellow River Basin and its coal mining areas holds significant importance for dynamic vegetation monitoring and mining area management. This study employs MODIS NDVI data and combines various analytical methods, including trend analysis and coefficient of variation analysis, to reveal the characteristics of NDVI spatiotemporal variations and their response to climatic factors in the study area. The results indicate the following: (1) The overall NDVI in the Shanxi section of the Yellow River Basin exhibits a growth trend with an annual growth rate of 1.82% and a 36% increase. Among the mining areas, the NDVI increase is most prominent in the Hebaopian mining area with a 100% growth, while the QinYuan mining area shows the lowest increase at 21%; (2) The NDVI in the Shanxi section of the Yellow River Basin displays high fluctuations, with areas of moderate and high fluctuations accounting for 54.39% of the total. The Hebaopian mining area has a substantial portion of high-fluctuation areas at 38.85%; (3) According to the Hurst index analysis, future vegetation changes in the Shanxi section of the Yellow River Basin are uncertain, with approximately 9.77% of areas expected to continue improving; (4) The variations in the NDVI and climatic factors across the Shanxi section of the Yellow River Basin display spatial heterogeneity. The NDVI exhibits a positive correlation with both temperature and precipitation, with the correlation with precipitation being more pronounced than that with temperature. Precipitation exerts a more significant influence on the NDVI than temperature. These findings not only provide scientific guidance for vegetation restoration and area management in the Shanxi section of the Yellow River Basin and its mining areas but also serve as a scientific basis for decision making regarding vegetation management under the influence of climate change and human activities.

A rapid and accurate determination of large-scale winter wheat yield is significant for food security and policy formulation. In this study, meteorological data and enhanced vegetation index (EVI) were used to estimate the winter wheat yield in Henan Province, China, by constructing a deep learning model. The deep learning model combines CNN feature extraction and makes full use of the sequence data processing capability of the LSTM and a multi-head attention mechanism to develop a novel CNN–MALSTM estimation model, which can capture the information of input sequences in different feature subspaces to enhance the expressiveness of the model. A CNN–LSTM baseline model was also constructed for comparison. Compared with the baseline model (R² = 0.75, RMSE = 646.53 kg/ha, and MAPE = 8.82%), the proposed CNN–MALSTM model (R² = 0.79, RMSE = 576.01 kg/ha, MAPE = 7.29%) could more accurately estimate the yield. Based on the cross-validation with one year of left-out data and the input of the fertility period by fertility period to explore the sensitivity of the model to data from different fertility periods to the final yield, an annual yield distribution map of Henan Province was constructed. Through cross-validation, the stability of the model in different years was assessed. The results showed that the model could obtain the best prediction of the yield approximately 20 days in advance. In terms of the spatial distribution of the yield in Henan Province on a yearly basis, the estimated yield showed an overall uptrend from west to east, consistent with the trend in the statistical yearbook of the yield for Henan Province. Thus, it can be concluded that the proposed CNN–MALSTM model can provide stable yield estimation results.