Water scarcity is a significant constraint in agricultural ecosystems of arid regions, necessitating sustainable development of agricultural water resources. This study innovatively combines Bayesian theory and Water Footprint (WF) to construct a Bayesian Network (BN). Water quantity and quality data were evaluated comprehensively by WF in agricultural production. This evaluation integrates WF and local water resources to establish a sustainability assessment framework. Selected nodes are incorporated into a BN and continuously updated through structural and parameter learning, resulting in a robust model. Results reveal a nearly threefold increase of WF in the arid regions of Northwest China from 1989 to 2019, averaging 189.95 × 10⁸ m³ annually. The region's agricultural scale is expanding, and economic development is rapid, but the unsustainability of agricultural water use is increasing. Blue WF predominates in this region, with cotton having the highest WF among crops. The BN indicates a 70.1 % probability of unsustainable water use. Sensitivity analysis identifies anthropogenic factors as primary drivers influencing water resource sustainability. Scenario analysis underscores the need to reduce WF production and increase agricultural water supply for sustainable development in arid regions. Proposed strategies include improving irrigation methods, implementing integrated water-fertilizer management, and selecting drought-resistant, economically viable crops to optimize crop planting structures and enhance water use efficiency in current agricultural practices in arid regions. This study not only offers insights into water management in arid regions but also provides practical guidance for similar agricultural contexts. The BN model serves as a flexible tool for informed decision-making in dynamic environments.

Estimating the fractional coverage of the photosynthetic vegetation (f_PV) and non-photosynthetic vegetation (f_NPV) is essential for assessing the growth conditions of vegetation growth in arid areas and for monitoring environmental changes and desertification. The aim of this study was to estimate the f_PV, f_NPV and the fractional coverage of the bare soil (f_BS) in the lower reaches of Tarim River quantitatively. The study acquired field data during September 2020 for obtaining the f_PV, f_NPV and f_BS. Firstly, six photosynthetic vegetation indices (PVIs) and six non-photosynthetic vegetation indices (NPVIs) were calculated from Sentinel-2A image data. The PVIs include normalized difference vegetation index (NDVI), ratio vegetation index (RVI), soil adjusted vegetation index (SAVI), modified soil adjusted vegetation index (MSAVI), reduced simple ratio index (RSR) and global environment monitoring index (GEMI). Meanwhile, normalized difference index (NDI), normalized difference tillage index (NDTI), normalized difference senescent vegetation index (NDSVI), soil tillage index (STI), shortwave infrared ratio (SWIR32) and dead fuel index (DFI) constitutes the NPVIs. We then established linear regression model of different PVIs and f_PV, and NPVIs and f_NPV, respectively. Finally, we applied the GEMI-DFI model to analyze the spatial and seasonal variation of f_PV and f_NPV in the study area in 2020. The results showed that the GEMI and f_PV revealed the best correlation coefficient (R²) of 0.59, while DFI and f_NPV had the best correlation of R² = 0.45. The accuracy of f_PV, f_NPV and f_BS based on the determined PVIs and NPVIs as calculated by GEMI-DFI model are 0.69, 0.58 and 0.43, respectively. The f_PV and f_NPV are consistent with the vegetation phonological development characteristics in the study area. The study concluded that the application of the GEMI-DFI model in the f_PV and f_NPV estimation was sufficiently significant for monitoring the spatial and seasonal variation of vegetation and its ecological functions in arid areas.

Snowfall is an important climatic variable with remarkable impacts on vegetation greenness in the alpine and extra-tropical regions. Central Asia (CA) is located in the middle latitude with high mountains, and the ecosystem is vulnerable to climate change in CA. In this region, the vegetation response to winter snowfall and its interactions with growing season climatic factors on vegetation greenness is still unclear. Thus, this study attempted to examine the impact of winter snowfall on vegetation greenness with remote-sensing vegetation index and reanalysis climatic data, and to investigate the interactions between winter snowfall and growing season climatic factors and their influence on vegetation growth via path analysis. The results showed that there is a generally positive correlation between winter snowfall and vegetation greenness from May to September and during the whole growing season (April to September). This positive correlation was significant in 8% of the study area for the whole growing season. However, the increase in winter snowfall was not beneficial to the regional vegetation growth in April. The vegetation response to winter snowfall also relates to land-cover type and elevation. The vegetation greenness in grassland was depicted to be more sensitive to winter snowfall than that in forestland. The response turned from positive to negative when the elevation increased from below 3000 m to above 3000 m. Moreover, the impact of winter snowfall on vegetation greenness was not regulated by temperature and precipitation in the growing season in more than 70% of CA. The impact relates to the interaction with April temperature in Central Kazakhstan, and is regulated by growing season precipitation in North Kazakhstan where annual precipitation mainly occurs in the growing season. The impact of winter snowfall on vegetation greenness is more important than growing season precipitation and temperature in some areas, since annual precipitation does not concentrate in the growing season or the relative increase of winter snowfall is great in these places. The results of the present study improve the understanding of vegetation response to climate change, and provide a scientific reference for environmental protection in CA and similar regions.

Recent studies on China’s arid and semi-arid regions, particularly the Tarim River Basin (TRB), have shown an increase in the intensity and frequency of extreme weather events. This research examines the link between meteorological droughts, as measured by the Standardized Precipitation Evapotranspiration Index (SPEI), and hydrological droughts, as indicated by the Standardized Runoff Index (SRI) and the Standardized Terrestrial Water Storage Index (STI), over various time scales. Historical data indicate that SPEI drought frequency (DF) was 14.3–21.9%, with prevalent events in the northern oases. SRI DF ranged from 9.0% to 35.8%, concentrated around the Taklamakan and Kumtag Deserts, while STI DF varied between 4.4% and 32.7%, averaging 15% basin-wide. Future projections show an increased DF of SPEI in deserts and a decrease in oases; SRI DF decreased in deserts but increased in oases. STI changes were more moderate. The study also found a higher risk of drought progression from SPEI to SRI in the southwestern and northeastern oases, exceeding 50% probability, while central and eastern TRB had lower risks. The western TRB and inner Taklamakan Desert faced higher risks of SPEI to STI progression, with probabilities over 45%, in contrast to the lower risks in the eastern and central oases. The concurrence of SRI/STI with moderate to extreme SPEI droughts led to a higher probability and area of SRI/STI droughts, whereas consistent SPEI types showed a reduced induced probability and extent of SRI/STI droughts. This study enhances the understanding of drought propagation from meteorological to hydrological droughts in the TRB and contributes to the prevention of hydrological drought to a certain extent.

Global warming is increasing the frequency and intensity of drought events, posing profound threats to ecosystems. Ecological drought is a multifaceted process in terrestrial ecosystems, characterized by the impairment of vegetation eco-physiological functions due to water stress. However, current ecological drought assessments primarily focus on ecological water supply and demand, often overlooking the status of vegetation growth. To address this gap, we developed a standardized composite ecological drought index (SCEDI) that integrates precipitation (Pre), ecological water deficit (EWD), and the vegetation health index (VHI) using a Euclidean distance-based approach. We further analyzed spatiotemporal dynamics of ecological drought and its characteristics, utilizing run theory and empirical orthogonal function (EOF) methods. Results showed that ecological drought intensified in spring (SCEDI change: -0.002) and winter (-0.013), but weakened in summer (0.006) and autumn (0.002). The first EOF mode explained 30.51 % of the total variance, indicating a reversal in wet-dry patterns between mountainous regions and oases around 2002. Over the past 34 years, Xinjiang has experienced more than 19 ecological drought events, each with an average duration of approximately 3.11 months. Regionally, droughts in mountainous areas occurred more frequently but were shorter and less intense, while in oasis areas they were less frequent but more persistent and severe. Furthermore, Shapley Additive Explanations (SHAP) analysis identified EWD (SHAP value: 0.265) and precipitation (0.176) as the most influential drivers of ecological drought. Higher values of both factors were positively associated with ecological drought severity, while lower values had a mitigating effect. This study highlights the potential of SCEDI in accurately capturing ecological drought dynamics and provides a scientific basis for ecological risk assessment and adaptive drought management in arid regions.

Grassland ecosystems in arid regions increasingly experience resilience loss due to intensifying climatic variability. However, the limited interpretability of conventional machine learning models constrains our understanding of underlying ecological drivers. This study constructs an integrative framework that combines temporal autocorrelation (TAC) metrics with explainable machine learning, employing Random Forest and SHAP (SHapley Additive exPlanations) analysis. Time series of satellite-derived vegetation indices from MODIS (2001–2023), particularly the kernel Normalized Difference Vegetation Index (KNDVI), support the generation of TAC and its trend-based derivative δTAC. The framework assesses ecosystem resilience across seven representative grassland types in Xinjiang, capturing diverse responses to climate variability and vegetation dynamics. Results reveal pronounced spatial heterogeneity: resilience declines in radiation-stressed arid zones, while hydrothermally stable regions maintain stronger recovery capacity. Key drivers include temperature variability and vegetation dynamics, with divergent effects among grassland types. Meadow and Typical Steppe exhibit higher resilience under stable hydrothermal regimes, whereas desert and alpine systems show greater sensitivity to warming and climatic fluctuations. This framework enhances diagnostic transparency and ecological insight, offering a spatially explicit, data-driven tool for resilience monitoring. The findings support the formulation of targeted adaptation strategies and sustainable grassland management in response to ongoing climate change.