As an essential link between terrestrial and climatic ecosystems, vegetation has been altered by the soil hy-drological environment associated with frozen soil thaw. However, it is not clear whether fluctuating soil moisture (SM) within the frozen soil zone alters the hydrologic environment to alleviate water stress in plants further, and there are scant previous studies at large scale on whether there is a threshold for SM on vegetation greening. This study integrated SM monitoring data at 125 stations from existing studies, then quantified the advantages of six remote sensing/reanalysis SM products: QTP-DNN-Sm, Global Land Data Assimilation System (GLDAS-Noah), European Centre for Medium-Range Weather Forecasts atmospheric reanalysis (ERA5-Land), European Space Agency Climate Change Initiative (ESA CCI), Global-SM, and QTP-SM. Moreover, we assessed the influence of single and multiple regional environmental elements (temperature, precipitation, land surface temperature (LST), normalized difference vegetation index (NDVI), and snow) on SM, as well as identified four trends of SM and vegetation growth for 51.26% of the Tibetan Plateau (TP). The results are as follows: 1) The overall performance of QTP-DNN-Sm products was slightly better than that of Global-SM, GLDAS-Noah, QTP-SM, ESA CCI, and ERA5-Land, with higher median Pearson correlation coefficient (R) value (0.685, 0.686, 0.699, 0.704, 0.300, and 0.582 for QTP-DNN-Sm, Global-SM, GLDAS-Noah, QTP-SM, ESA CCI, and ERA5-Land, respectively) and lowest median unbiased Root Mean Square Error (ubRMSE) (0.061, 0.064, 0.068, 0.064, 0.042, 0.076, and 0.047 m3/m3 for QTP-DNN-Sm, Global-SM, GLDAS-Noah, QTP-SM, ESA CCI, ERA5-Land, and ESA CCI, respectively). 2) NDVI in the frozen soil zone was the best variable to explain SM based on the GeoDetector-based factor detection, and interaction detection results indicated that the interaction between NDVI and temperature was gradually emerging to explain SM from permafrost zones to seasonally frozen ground zones. 3) The nonlinear relationship function between SM and NDVI showed that vegetation growth in 47.76% of the area (mainly distributed in the Changjiang River, Yarlung Zangbo River, and Yellow River basins) was more influenced by phenology. Thresholds existed in 3.49% of TP, where the cumulative effect of SM affects vegetation growth. In 0.65% of the regions, vegetation growth experienced eco-physiological processes of positive relief of water stress and physical processes of negative damage. The ease with which SM altered vegetation growth trends was consistent with the degradation degree of frozen soil type. Although the percentage of regions where the thresholds exist is relatively small, the positive/negative effects of the complex localized inter between SM and vegetation in these regions could threaten the balance and stability of fragile alpine ecosystems sustained by permafrost.

Aboveground biomass (AGB) serves as a crucial measure of ecosystem productivity and carbon storage in alpine grasslands, playing a pivotal role in understanding the dynamics of the carbon cycle and the impacts of climate change on the Qinghai-Xizang Plateau. This study utilized Google Earth Engine to amalgamate Landsat 8 and Sentinel-2 satellite imagery and applied the Random Forest algorithm to estimate the spatial distribution of AGB in the alpine grasslands of the Beiliu River Basin in the Qinghai-Xizang Plateau permafrost zone during the 2022 growing season. Additionally, the geodetector technique was employed to identify the primary drivers of AGB distribution. The results indicated that the random forest model, which incorporated the normalized vegetation index (NDVI), the enhanced vegetation index (EVI), the soil-adjusted vegetation index (SAVI), and the normalized burn ratio index (NBR2), demonstrated robust performance in regards to AGB estimation, achieving an average coefficient of determination (R2) of 0.76 and a root mean square error (RMSE) of 70 g/m2. The average AGB for alpine meadows was determined to be 285 g/m2, while for alpine steppes, it was 204 g/m2, both surpassing the regional averages in the Qinghai-Xizang Plateau. The spatial pattern of AGB was primarily driven by grassland type and soil moisture, with q-values of 0.63 and 0.52, and the active layer thickness (ALT) also played a important role in AGB change, with a q-value of 0.38, demonstrating that the influences of ALT should not be neglected in regards to grassland change.

Surface albedo is a quantitative indicator for land surface processes and climate modeling, and plays an important role in surface radiation balance and climate change. In this study, by means of the MCD43A3 surface albedo product developed on the basis of Moderate Resolution Imaging Spectroradiometer (MODIS), we analyzed the spatiotemporal variation, persistence status, land cover type differences, and annual and seasonal differences of surface albedo, as well as the relationship between surface albedo and various influencing factors (including Normalized Difference Snow Index (NDSI), precipitation, Normalized Difference Vegetation Index (NDVI), land surface temperature, soil moisture, air temperature, and digital elevation model (DEM)) in the north of Xinjiang Uygur Autonomous Region (northern Xinjiang) of Northwest China from 2010 to 2020 based on the unary linear regression, Hurst index, and Pearson's correlation coefficient analyses. Combined with the random forest (RF) model and geographical detector (Geodetector), the importance of the above-mentioned influencing factors as well as their interactions on surface albedo were quantitatively evaluated. The results showed that the seasonal average surface albedo in northern Xinjiang was the highest in winter and the lowest in summer. The annual average surface albedo from 2010 to 2020 was high in the west and north and low in the east and south, showing a weak decreasing trend and a small and stable overall variation. Land cover types had a significant impact on the variation of surface albedo. The annual average surface albedo in most regions of northern Xinjiang was positively correlated with NDSI and precipitation, and negatively correlated with NDVI, land surface temperature, soil moisture, and air temperature. In addition, the correlations between surface albedo and various influencing factors showed significant differences for different land cover types and in different seasons. To be specific, NDSI had the largest influence on surface albedo, followed by precipitation, land surface temperature, and soil moisture; whereas NDVI, air temperature, and DEM showed relatively weak influences. However, the interactions of any two influencing factors on surface albedo were enhanced, especially the interaction of air temperature and DEM. NDVI showed a nonlinear enhancement of influence on surface albedo when interacted with land surface temperature or precipitation, with an explanatory power greater than 92.00%. This study has a guiding significance in correctly understanding the land-atmosphere interactions in northern Xinjiang and improving the regional land-surface process simulation and climate prediction.

Active layer thickness (ALT) of permafrost changes significantly under the combined influence of human activities and climate warming, which has a significant impact on the ecological environment, hydrology, and engineering construction in cold regions. The spatial differentiation of Active layer thickness and its influencing factors have become one of the hot topics in the field of cryopedology in recent years, but there are few studies in the Da Hinggan Ling Prefecture (DHLP). In this study, the Stefan equation was used to simulate the Active layer thickness in the Da Hinggan Ling Prefecture, and the factor detection and interaction detection functions of geodetector were used to analyze the factors affecting the spatial differentiation of Active layer thickness from both natural and humanity aspects. The results showed that Active layer thickness in the Da Hinggan Ling Prefecture ranges from 58.82 cm to 212.55 cm, the determinant coefficient R-2, MAE, RMSE between simulation results and the sampling points data were 0.86, 11.25 (cm) and 13.25 (cm), respectively. Lower Active layer thickness values are mainly distributed higher elevations in the west, which are dominated by forest (average ALT: 136.94 cm) and wetlands (average ALT: 71.88 cm), while the higher values are distributed on cultivated land (average ALT: 170.35 cm) and construction land (average ALT: 176.49 cm) in the southeast. Among the influencing factors, elevation is significantly negatively correlated with ALT. followed by summer mean LST, SLHF and snow depth. NDVI and SM has the strong explanation power for the spatial differentiation of ALT in factor detection. Regarding interactions, the explanatory power of slope boolean AND snow depth is the highest of 0.83, followed by the elevation boolean AND distance to settlements. The results can provide reference for the formulation of ecological environmental protection and engineering construction policies in cold regions.

Kazakhstan is part of the Eurasian Steppes, the world's largest contiguous grassland system. Kazakh grassland systems are largely understudied despite being historically important for agropastoral practices. These grasslands are considered vulnerable to anthropogenic activities and climatic variability. Few studies have examined vegetation dynamics in Central Asia owing to the complex impacts of moisture, climatic and anthropogenic forcings. A comprehensive analysis of spatiotemporal changes of vegetation and its driving factors will help elucidate the causes of grassland degradation. Here, we investigated the individual and pairwise interactive influences of various social-environmental system (SES) drivers on greenness dynamics in Kazakhstan. We sought to examine whether there is a relationship between peak season greenness and its drivers - spring drought, preceding winter freeze-thaw cycles, percent snow cover and snow depth - for Kazakhstan during 2000-2016. As hypothesized, snow depth and spring drought accounted for 60 % and 52 % of the variance in the satellite-derived normalized difference vegetation index (NDVI) in Kazakhstan. The freeze-thaw process accounted for 50 % of NDVI variance across the country. In addition, continuous thawing during the winter increased vegetation greenness. We also found that moisture and topographic factors impacted NDVI more significantly than socioeconomic factors. However, the impacts of socioeconomic drivers on vegetation growth were amplified when they interacted with environmental drivers. Terrain slope and soil moisture had the highest q-values or power of determinant, accounting for -70 % of the variance in NDVI across the country. Socioeconomic drivers, such as crop production (59 %), population density (48 %), and livestock density (26 %), had significant impacts on vegetation dynamics in Kazakhstan. We found that most of the pairwise interactive influences of the drivers exhibited bi-factor enhancement, and the interaction between soil moisture and elevation was the largest (q = 0.92). Our study revealed the optimal ranges and tipping points of SES drivers and quantified the impacts of various driving factors on NDVI. These findings can help us identify the factors causing grassland degradation and provide a scientific basis for ecological protection in semiarid regions.

The glacier is a crucial freshwater resource in arid and semiarid regions, and the vulnerability of the glacier change is intimately linked to regional ecological services and socio-economic sustainability. Taking the Tianshan Mountains region in China as an example, a basic framework for studying the vulnerability of glacier change was constructed so as to address factors such as physical geography, population status, socio-economic level, agricultural development, and social services. The framework was based on key dimensions, that is, exposure, sensitivity, and adaptability, and this constituted a targeted evaluation index system. We examined the spatial structure and spatial autocorrelation of the glacier change vulnerability using ArcGIS and GeoDa software. The influence and interaction of natural, social, economic, population and other factors on glacier change adaptability was examined using the GeoDetector model. The results suggested the following: (1) The vulnerability level decreased from the western region to the eastern region with significant differences between the two regions. The eastern region had the lowest vulnerability, followed by the central region, and then western region which had the highest vulnerability. (2) Significant positive and negative correlations were found between exposure, sensitivity, and adaptability, indicating that the areas with high exposure and high sensitivity to glacier change tended to have a low adaptive capacity, which led to high vulnerability, and vice versa. (3) The spatial heterogeneity regarding the ability to cope with glacier change reflected the combined effects of the natural, social, economic, and demographic factors. Among them, factors such as the production value of secondary and tertiary industries, the urban population, urban fixed-asset investment, and the number of employees played major roles regarding the spatial heterogeneity of glacier change.