Soil moisture (SM) is a critical parameter in Earth’s water cycle, significantly impacting hydrological, agricultural, and meteorological research fields. The challenge of estimating surface soil moisture from synthetic aperture radar (SAR) data is compounded by the influence of vegetation coverage. This study focuses on the Weigan River and Kuche River Delta Oasis in Xinjiang, employing high-resolution Sentinel-1 and Sentinel-2 images in conjunction with a modified Water Cloud Model (WCM) and the grayscale co-occurrence matrix (GLCM) for feature parameter extraction. A soil moisture inversion method based on stacked ensemble learning is proposed, which integrates random forest, CatBoost, and LightGBM. The findings underscore the feasibility of using multi-source remote sensing data for oasis moisture inversion in arid regions. However, soil moisture content estimates tend to be overestimated above 10% and underestimated below 5%. The CatBoost model achieved the highest accuracy (R² = 0.827, RMSE = 0.014 g/g) using the top 16 feature parameter groups. Additionally, the R² values for Stacking1 and Stacking2 models saw increases of 0.008 and 0.016, respectively. Thus, integrating multi-source remote sensing data with Stacking models offers valuable support and reference for large-scale estimation of surface soil moisture content in arid oasis areas.

Soil salinization is a significant global threat to sustainable agricultural development, with soil salinity serving as a crucial indicator for evaluating soil salinization. Remote sensing technology enables large-scale inversion of soil salinity, facilitating the monitoring and assessment of soil salinization levels, thus supporting the prevention and management of soil salinization. This study employs multi-source remote sensing data, selecting 8 radar polarization combinations, 10 spectral indices, and 3 topographic factors to form a feature variable dataset. By applying a normalized weighted variable optimization method, highly important feature variables are identified. AdaBoost, LightGBM, and CatBoost machine learning methods are then used to develop soil salinity inversion models and evaluate their performance. The results indicate the following: (1) There is generally a strong correlation between radar polarization combinations and vegetation indices, and a very high correlation between various vegetation indices and the salinity index S3. (2) The top five feature variables, in order of importance, are Aspect, VH², Normalized Difference Moisture Index (NDMI), VH, and Vegetation Moisture Index (VMI). (3) The method of normalized weighted importance scoring effectively screens important variables, reducing the number of input feature variables while enhancing the model’s inversion accuracy. (4) Among the three machine learning models, CatBoost performs best overall in soil salt content (SSC) prediction. Combined with the top five feature variables, CatBoost achieves the highest prediction accuracy (R² = 0.831, RMSE = 2.653, MAE = 1.034) in the prediction phase. This study provides insights for the further development and application of methods for collaborative inversion of soil salinity using multi-source remote sensing data.

Ecosystems offer natural resources and habitats for humans, serving as the foundation for human social development. Taking the Tianshan Mountains as the study area, this study investigated the changing trends, hot spots, and driving factors of water yield (WY), soil conservation (SC), carbon storage (CS), and habitat quality (HQ), in the Tianshan region, from 1990 to 2020. To determine the trade-offs and synergies between the ESs, we employed the Spearman correlation coefficient, geographically weighted regression, the self-organizing map (SOM), and other methods. Five main results were obtained. (1) There were similar spatial distribution patterns for WY, HQ, CS, and SC, with high-value areas mainly concentrated in grassland zones, forest zones, river valleys, and the intermountain basins of the mountain range, while regions with low value were clustered in desert zones and snow/ice zones. (2) According to the hotspot analysis, areas with relatively strong ES provisioning for WY, HQ, CS, and SC, were primarily concentrated in the BoroHoro Ula Mountains and Yilianhabierga Mountains. In contrast, areas with relatively weak ES provisioning were mainly located in the Turpan Basin. (3) Precipitation was the primary explanatory factor for WY. Soil type, potential evapotranspiration (PET), and the normalized difference vegetation index (NDVI) were the primary explanatory factors for HQ. Soil type and NDVI were the primary explanatory factors for CS. PET was the primary explanatory factor for SC. (4) There were synergistic relationships between the WY, HQ, CS, and SC, with the strongest synergies found between CS–HQ, WY–HQ, and WY–SC. (5) Six ES bundles were identified through the SOM method, with their composition varying at different spatial scales, indicating the need for different ES management priorities in different regions. Our analysis of ESs, from various perspectives, offers insights to aid sustainable ecosystem management and conservation efforts in the Tianshan region and other major economic areas worldwide.

Arid region ecosystems are among the most fragile ecological types worldwide. They depend heavily on limited water resources and are strongly influenced by intensive human activities, leading their ecosystem services to exhibit nonlinear and threshold responses to driving factors. Identifying the thresholds of ecosystem services under the combined influence of natural and socio-economic interactive drivers is of great significance for regional ecological risk warning and differentiated management. Taking the Tarim River Basin as a case study, this research establishes an integrated analytical framework that combines causal inference, interaction term construction, interpretable machine learning (XGBoost-SHAP), and piecewise linear regression. The framework is used to evaluate the variations in four types of ecosystem services in 2000, 2010, and 2023, to analyze the interactive effects of driving factors, and to identify their thresholds influencing ecosystem service functions. The results indicate that (1) different types of ecosystem service functions exhibited distinct trends from 2000 to 2023, with habitat quality and water yield showing declining tendencies, while soil conservation and Windbreak and sand fixation demonstrated gradual increases; (2) Causal Screening and interaction modeling revealed that the interaction between precipitation and population density (Pre × Pop) served as the key synergistic driver of changes in the four ecosystem service functions. Both the ecosystem services and the coupled natural–social driving processes exhibited pronounced nonlinear characteristics, with evident trend shifts occurring within specific threshold intervals. (3) The precise coupling thresholds of different ecosystem services under natural–social drivers were identified, intuitively revealing the coupling threshold characteristics of various ecosystem services; (4) The integration of causal inference with interpretable machine learning enhances the reliability of threshold identification, revealing the heterogeneous response mechanisms of different services and providing a quantitative basis for the zoning regulation and differentiated management of regional ecosystems. The findings offer a transferable methodological framework to support ecological governance in arid regions.

In arid regions, potentially toxic elements (PTEs) can accumulate in oasis farmland soils, posing risks to both ecosystems and human health and threatening the long-term sustainability of agricultural production. However, we still lack a clear understanding of how multiple elements co-accumulate and what non-linear processes drive their buildup. Here, we investigated typical agricultural soils in the Aksu area of Xinjiang. We measured 12 elements (As, Cr, Cu, Ni, Zn, Co, V, Se, F, Ba, Sn, Mn) in 28 surface samples. To assess the pollution levels, we used three indices: the single-factor index (P_i), the geo-accumulation index (I_geo), and the Nemerow composite index. Source apportionment was performed with positive matrix factorization (PMF). We then built an XGBoost model to predict the Nemerow index, and applied SHAP (Shapley additive explanations) to quantify the marginal contribution and non-linear response of each element. Our results show that the average concentrations of Se, F, and As are 1.47, 1.27, and 1.35 times the national background values, respectively. The exceedance rate (P_i > 1) for these elements ranges from 78.6% to 92.9%. Nevertheless, the overall pollution is mild: only one out of 28 sampling sites (3.6%) falls into the moderately polluted category. PMF resolved three major sources: (1) parent material plus evaporation enrichment (F, Mn, and Ba, ~45% of the total contribution); (2) agricultural and anthropogenic activities (As, Cr, V, Zn, ~40%); and (3) local industrial or waste inputs (Sn and Ba, ~15%). The XGBoost model shows good predictive performance on the test set (R² = 0.864, RMSE = 0.104). SHAP analysis reveals that Se, F, and As are the main drivers of the composite pollution index. Se has a clear threshold: once its concentration goes above 0.3 mg·kg⁻¹, its positive contribution jumps sharply. Overall, the farmland soils in Aksu show mild enrichment of several elements, with Se and F as the main indicators. Evaporation enrichment and farming practices are the dominant processes behind this enrichment. The integrated framework—pollution indices, PMF source apportionment, XGBoost prediction, and SHAP interpretation—provides a scientifically sound way to manage soil environments in arid regions.

Long-term ecological water conveyance has been implemented to reconnect hyper-arid inland-river corridors, but ecological effectiveness cannot be inferred from delivered water volume alone. We assessed ecohydrological changes in the lower Tarim River from 2000 to 2023 using annual water conveyance records, groundwater observations, growing season Carnegie–Ames–Stanford Approach (CASA) net primary productivity (NPP), Landsat-derived Normalized Difference Salinity Index (NDSI), climate data, and four operational NPP-NDSI states. Annual conveyance averaged 4.08 × 10⁸ m³ and peaked in 2017 at 12.09 × 10⁸ m³. Mean groundwater depth decreased from 7.49 m in 2010 to 4.25 m in 2023. Current-inclusive multi-year conveyance windows were more strongly negatively associated with groundwater depth than same-year conveyance (W3: rho = −0.743; W4: rho = −0.782); Newey–West tests remained significant at lags 1–3, and moving-block bootstrap intervals excluded zero for W3 and W4, whereas the strictly antecedent three-year metric was less robust. The high-NDSI area decreased from 1795.04 to 42.19 km², whereas the higher-productivity area increased from 91.81 to 1127.81 km². A paired Enhanced Thematic Mapper Plus/Operational Land Imager (ETM+/OLI) analysis showed high NDSI correspondence (r = 0.980) but a mean OLI-minus-ETM+ bias of −0.0122; the corresponding 2010 cross-sensor adjustment was 63.9% of the absolute raw 2010–2015 mean-NDSI shift. Cross-era NDSI class areas and sampled-state trajectories are therefore treated as sensor-sensitive descriptive quantities. Spatial support sensitivity was also important: the functionally limited low-salinity state (FLLS) was the largest 2023 state on the fine-alignment grid (53.93%), whereas the higher-productivity low/moderate-NDSI state (HPLN) became dominant after aggregation to 500 m (66.57%); so, no scale-invariant claim is made about the dominant 2023 state. A fixed-2000-land-cover CASA run produced no material change in NPP because the land-cover transitions represented in the common mask occurred among classes sharing the same NDVI endmembers and εmax. Overall, the results support asynchronous indicator change and a state-based monitoring framework, while temporal dependence, irregular groundwater sampling, residual spatial autocorrelation, cross-sensor discontinuity, and spatial support constrain causal and fine-scale inference.