Measuring the impacts of anthropogenic activities on grassland ecosystems is of great significance to establish effective grassland management policies. Due to the uncertainty and complexity of the anthropogenic activities, the impacts are difficult to quantify, especially at large spatial and long temporal scales. We used the Zhou Guangsheng model to simulate climate‐driven potential net primary productivity (NPP_P) and MOD17A3 NPP data to acquire actual NPP (NPP_A) influenced by both anthropogenic activities and climate change over the temperate grassland of Inner Mongolian from 2000 to 2014. We then simulated human‐induced NPP (NPP_H), calculated as the difference between actual and potential NPP (NPP_H = NPP_A − NPP_P), to determine anthropogenic effects on the temperate grassland. By comparing the NPP_H and its time‐series linear slope from 2000 to 2014, we further proposed six types of anthropogenic impacts. Along the temperate grassland climate gradient, vegetation changed from meadow steppe in the east to typical steppe in the middle and desert steppe in the west, and NPP_A and NPP_P decreased. NPP_H had no obvious spatial pattern. We observed positive effect (NPP_H > 0) on only 4.2% of total grassland area. The area where the types of “positive effect (NPP_H > 0) decrease – negative effect (NPP_H < 0) increase” and “negative effect (NPP_H < 0) increase” were observed in time‐series trends accounted for 68.8% of total grassland area. The grazing withdrawal and ecological compensation programs implemented around 2004 did not lead to a decline in the number of livestock and NPP_H. Governments need to improve livestock number controls and strengthen grazing regulations.

The Tibetan Plateau (TP) is crucial for global climate change and China’s ecological security. Given recent drastic changes in vegetation from climate change and human activities, long-term vegetation monitoring is urgently required. This study produced the vegetation maps of the TP from 1990 to 2020 every ten years using random forest classifier and Landsat imagery. We selected the same stable samples and features for mapping to reduce errors between years and proposed spatial filtering to further improve the accuracy. The overall accuracy surpassed 95.00%, with all Kappa coefficients exceeding 0.95. A further assessment based on sampling sites from literature and field survey was higher than 80%. The importance ranking results indicated that in the TP, climate factors and terrain factors are the most important factors in the vegetation mapping. This study provides a method for mapping vegetation in alpine areas and data support for researching the dynamic change of vegetation on the TP and evaluating its response to climate change.

Phase unwrapping is one of the key processing steps of D-InSAR (Differential Interferometric Synthetic Aperture Radar), and its accuracy has a great impact on the accuracy of interferometric deformation results. Therefore, it is necessary to choose the proper phase unwrapping method for mining deformation monitoring with different coherence by the comparative study of different phase unwrapping methods, because the coherence is different for different mining areas. In this paper, 2 data sets of different coherence Sentinel-1A data in Xuangang mining areas were processed using 3 different D-InSAR unwrapping methods: Region Growing, Minimum Cost Flow, and Delaunay MCF by SARscape software. Through statistical analysis of phase unwrapping results and comparative analysis of the difference between the theoretical deformation value and the deformation value calculated by SARscape, the Delaunay MCF unwrapping method is found to be the best to monitor low-coherence mining areas under most conditions. However, the Minimum Cost Flow method is a better choice when the coherence of mining areas is high, and smoother and more continuous phase unwrapping results can be obtained by selecting a mining area with better coherence.

Litter has a special ecological functioningin grasslands. Few studies have been conducted to estimate litter mass using remotely sensed data during nongrowing seasons in arid grasslands although it is important forage for livestock sustainability. With MODIS data, estimation methods were developed for litter mass in the desert steppe of Inner Mongolia calibrated with field surveys. As MODIS Band 7 is located in the lignocellulose absorption pit of litter near 2100 nm, the best models were obtained for NDTI (normalized difference tillage index) (normalized difference between Bands 6 and 7) and STI (soil tillage index) (ratio of Band 6–7) among soil-unadjusted indices, and for MSACRI (modified soil-adjusted crop residue index) (modification of NDTI by incorporating soil line) among soil-adjusted indices. NDTI and STI explained 63% of the variance of litter mass, while MSACRI explained 71% of the variance. If data are not available for calculating soil line, it may be appropriate to use the soil-adjusted NDTI (S-NDTI), a new index proposed in the study that incorporates a soil adjustment factor into the NDTI equation. The optimal S-NDTI explained 66% of the variance. The NDTI, STI, MSACRI and S-NDTI can be applied to estimate litter mass in arid grasslands.

Accurate remote sensing identification of river ice not only provides scientific evidence for climate change but also offers early warning information for disasters such as ice jams. Currently, many researchers have used remote sensing index-based methods to identify river ice in alpine regions. However, in high-altitude areas, these index-based methods face limitations in recognizing river ice and distinguishing ice-snow mixtures. With the rapid advancement of machine learning techniques, some scholars have begun to use machine learning methods to extract river ice in northern latitudes. However, there is still a lack of systematic studies on the ability of machine learning to enhance river ice identification in high-altitude, complex terrains. The study evaluates the performance of machine learning methods and the RDRI index method across six aspects: river type, altitude, river width, ice periods, satellite data, and snow cover interference. The results show that machine learning, particularly the RF method, demonstrates superior generalization ability and higher recognition accuracy for river ice in the complex high-altitude terrain of the Tibetan Plateau by leveraging a variety of input data, including spectral and topographical information. The RF model performs best under all types of test conditions, with an average Kappa coefficient of 0.9088, outperforming other machine learning methods and significantly outperforming the traditional exponential method, demonstrating stronger recognition capabilities. Machine learning methods are adaptable to different types of river ice, showing particularly improved recognition of river ice in braided river systems. RF and SVM exhibit more accurate river ice recognition across different altitudinal gradients, with RF and SVM significantly improving the identification accuracy of river ice (0–90 m) on the plateau. RF and SVM methods offer more precise boundary recognition when identifying river ice across different ice periods. Additionally, RF demonstrates better generalization in the transfer of multisource satellite data. RF’s performance is outstanding under different snow cover conditions, overcoming the limitations of traditional methods in identifying river ice under thick snow. Machine learning methods, which are well suited for large sample learning and have strong generalization capabilities, show significant potential for application in river ice identification within high-altitude, complex terrains.

Under the ongoing trend of climate warming and increasing humidity on the Qinghai–Tibet Plateau, the Three River Source Region (TRSR) has shown strong sensitivity to global climate change. Its vegetation change is particularly worthy of attention and research. The Normalized Difference Vegetation Index (NDVI) is a key indicator for assessing the growth status of vegetation. However, the insufficiency of existing NDVI datasets in terms of spatiotemporal continuity has limited the accuracy of long-term vegetation change studies. This study proposed a machine learning-based downscaling framework that integrates the Moderate-resolution Imaging Spectroradiometer (MODIS) NDVI and the Global Inventory Monitoring and Modeling System (GIMMS) NDVI data to reconstruct a long-term, high-resolution NDVI dataset. Unlike conventional statistical fusion approaches, the proposed framework employs machine learning-based nonlinear relationships to generate long-term, high-resolution NDVI data. Three machine learning algorithms—Random Forest (RF), LightGBM, and CatBoost—were evaluated. Their performance was validated using the MODIS NDVI as reference, with the coefficient of determination (R²), root mean square error (RMSE), mean absolute error (MAE), and Pearson’s correlation coefficient (R) as evaluation metrics. Based on model comparison, the CatBoost model was identified as the optimal algorithm for spatiotemporal data fusion (R² = 0.9014, RMSE = 0.0674, MAE = 0.0445), significantly outperforming RF and LightGBM models and demonstrating stronger capability for NDVI spatiotemporal reconstruction. Using this model, a long-term, 1 km monthly GIMMS-MODIS NDVI dataset from 1982 to 2014 was successfully reconstructed. On the basis of this dataset, the spatiotemporal variation characteristics of vegetation in the TRSR from 1982 to 2014 were systematically analyzed. The research results show that: (1) The constructed long-series high-resolution NDVI dataset has a high consistency with MODIS NDVI data; (2) From 1982 to 2014, the NDVI in the TRSR showed an increasing trend, with an average growth rate of 0.0020/10a (p < 0.05). NDVI showed obvious spatial heterogeneity, characterized by a decreasing gradient from southeast to northwest. (3) The Yellow River source exhibited the most evident vegetation recovery, the Yangtze River Source area showed a moderate improvement, whereas the Lancang River Source area displayed little noticeable change. (4) Broad-leaved forests experienced the most significant growth, while cultivated vegetation displayed a marked tendency toward degradation. This study provides both a high-accuracy long-term NDVI product for the TRSR and a methodological foundation for advancing vegetation dynamics research in other high-altitude regions.