Fixed-depth (FD) assumptions can substantially bias estimates of soil water storage (SWS) integrated over the 0–200 cm profile, hereafter referred to as shallow SWS, particularly in mountain regions where soil thickness (ST) varies markedly and shallow bedrock is widespread. In this study, we developed an ST-constrained framework to reconstruct monthly shallow SWS across the Tibetan Plateau (TP) at 1 km resolution over an 11-year window comprising 132 months (January 2009–December 2019) by integrating remote-sensing products, reanalysis data, and field observations. Using 1,615 in situ ST profiles and multi-source environmental covariates, we first mapped ST with a random forest model and then downscaled GLDAS soil moisture (SM) for four standard layers: 0–10, 10–40, 40–100, and 100–200 cm. Pixel-level SWS was estimated by truncating the vertical SM profile at the predicted ST, thereby preventing water from being allocated below the soil–bedrock interface. In spatial cross-validation, the ST model showed moderate spatial transferability (R 2 = 0.46, CCC = 0.62, and ubRMSE = 41.78 cm). Downscaled SM in the 0–100 cm profile was evaluated against Tibet-Obs measurements for the three upper layers. Station-based temporal Pearson’s R (mean R 2; CCC) values were 0.56 (0.40; 0.23), 0.28 (0.24; 0.06), and 0.28 (0.16; 0.02) for the 0–10, 10–40, and 40–100 cm layers, respectively. The corresponding ubRMSE (bias) values were 0.15 (−0.03), 0.12 (+0.04), and 0.15 (+0.03) m3/m3, respectively. Point-scale performance was comparatively stronger in the surface layer but weak in the 40–100 cm layer (R = 0.28; CCC = 0.02); therefore, layer-resolved SWS estimates in deep-soil regions should be interpreted cautiously as predictor-conditioned downscaled reanalysis rather than as fully in situ-validated estimates. Direct validation of the 100–200 cm layer was not feasible because in situ observations below 80 cm were unavailable; therefore, this layer should be interpreted as a profile-consistent reconstruction rather than a directly validated product. Because 81.30% of the estimated difference between the FD and ST-constrained approaches originated from the reconstructed 100–200 cm layer, the absolute magnitude of the reported storage difference and trend reduction should be interpreted with caution, although the direction of the structural bias was consistent across GLDAS and ERA5-Land. The ST-constrained SWS averaged 504.06 km3, whereas the conventional FD approach yielded 1,255.85 km3. Thus, ignoring soil–bedrock constraints caused an overestimate of 751.79 km3 (149.15%). Sensitivity analysis further showed that FD-based SWS increased nearly linearly with assumed profile depth, whereas ST-constrained estimates stabilized once the physically available storage capacity was reached. The reconstructed dataset captured a pronounced seasonal cycle and a significant wetting trend. Specifically, the ST-constrained framework indicated a moderate increase of 4.99 km3 yr−1 during 2009–2019, compared with 10.78 km3 yr−1 under the FD assumption. These findings show the importance of incorporating ST as a spatially explicit lower boundary when reconstructing shallow subsurface water storage in topographically heterogeneous mountain regions.

Understanding the spatial distribution of soil organic carbon (SOC) on the Tibetan Plateau (TP) and its response to future climate change is crucial for regional carbon cycling and ecosystem management. This study used 372 soil profiles, including data from a dedicated field survey, to model SOC density (SOCD) at depths of 0-30, 30-50, and 50-100 cm using a recursive feature elimination-random forest (RFE-RF) approach. Based on 10-fold cross-validation, the model explained 56%, 48%, and 35% of the spatial variability in SOCD across these layers. For the baseline period (1990-2025), the total SOC stock in the upper 1 m was estimated at 34.17 Pg (90% prediction interval: 6.29-90.25 Pg). Of this total, 16.29 Pg (4.16-36.48 Pg), 6.85 Pg (0.84-18.75 Pg), and 11.03 Pg (1.29-35.02 Pg) were stored in the 0-30, 30-50, and 50-100 cm layers, respectively. Shapley additive explanations (SHAP) analysis identified the aridity index (AI) as the most influential driver, with higher aridity consistently exerting negative effects on SOCD across all depths. In the topsoil (0-30 cm), additional key drivers included the normalized difference vegetation index (NDVI), soil pH, and mean annual precipitation (MAP). The influence of potential evapotranspiration (PET) increased in the 30-50 cm layer, whereas silt content and shortwave infrared 1 (SWIR1) reflectance became more influential in the deepest layer (50-100 cm). Projections under the SSP5-8.5 scenario suggest a significant decline (p < 0.01) in SOC stocks from 2030 to 2100, particularly in the 50-100 cm layer. While humid to semi-arid regions experienced substantial SOC losses, the ariddesert region showed a significant increase in surface SOC (0-30 cm). These results clarify layer-specific drivers and future vulnerabilities of SOC on the TP and provide useful information for adaptive soil carbon management

Soil organic carbon (SOC) stocks within the active layer of the Tibetan Plateau (TP) permafrost are highly sensitive to climate change. This study employs a data-driven approach to map interannual variations in active layer thickness (ALT) and SOC density across the TP permafrost from 2016 to 2023, addressing limitations of previous static assessments. Using equal-area quadratic smoothing splines for pixel-level SOC estimation, we achieved precise stock calculations and substantially reduced uncertainties. Compared to the 2006–2015 baseline, the average ALT increased by approximately 2 cm during 2016–2023, with pronounced warming-driven deepening from 2019 to 2023 (p < 0.05). SOC stocks in the active layer also increased significantly over 2019–2023. Dynamic analyses indicate that carbon stored at 6–10 m depth may be released even when ALT remains below 6 m. Overall, the TP permafrost experienced a net loss of 0.49 Pg SOC at an average rate of 0.07 Pg yr−1, with accelerated carbon release observed in the Qiangtang Plateau under warming and drying conditions. These results enhance understanding of SOC dynamics, the interaction between ALT and SOC, and carbon–climate feedback mechanisms in the TP permafrost region.

Quantifying the dynamics of above and belowground biomass carbon (AGBC and BGBC) is essential for optimizing carbon sink management. However, the environmental thresholds that govern these dynamics under climate change remain poorly understood in China. In this study, we identified key thresholds by examining the relationships between AGBC (4485 observations) and BGBC (3442 observations) with mean annual temperature (MAT), aridity index (AI), and soil pH. Thresholds for AGBC were 15.24 degrees C (MAT), 1.17 (AI), and 6.87 (pH), while those for BGBC were 14.37 degrees C, 0.65, and 7.99, respectively. Additionally, we explored these thresholds in different ecosystems (forests, grasslands, shrublands, and wetlands). By spatially mapping these thresholds, we delineated environmentally sensitive areas-regions currently below (or above) the thresholds that are projected to exceed (or fall below) them under future climate scenarios. Using machine learning algorithms, we modeled AGBC and BGBC distributions for the years 2010 and 2100 (SSP5-8.5 scenario) and identified regions with the most significant expected changes. Overlaying threshold-sensitive areas with projected vegetation carbon changes revealed that AGBC is likely to increase in the southeastern Tibetan Plateau, while BGBC is projected to increase in the northern Shandong Province (Likelihood > 66 %). These shifts are primarily driven by regional warming and humidification that exceed identified MAT and AI thresholds. By integrating threshold identification with spatial and temporal analyses, this study enhances our understanding of vegetation carbon responses to climate change.

Accurate assessment of soil organic carbon (SOC) stocks is essential for understanding the global carbon cycle. However, substantial uncertainties remain in mountainous regions such as Tibet, where the commonly used fixed-depth (FD) approach overlooks spatial variability in soil thickness (ST). This omission leads to systematic overestimation in shallow soils (due to the accounting of carbon in non-existent soil layers) and underestimation in deep, incompletely sampled profiles. To address these dual biases, this study introduces an integrated framework that, for the first time, couples the equivalent soil mass (ESM) method—which corrects for bulk density–related biases—with spatial modeling constrained by ST. The methodology involved: (1) deriving point-scale ESM-based SOC stocks for standard depth intervals (0–30, 30–50, and 50–100 cm) using equal-area quadratic smoothing splines; (2) generating spatially explicit, wall-to-wall predictions at 1-km resolution through quantile regression forests, employing an inverse probability of censoring weighting QRF (IPCW-QRF) for ST and a standard QRF for SOC; and (3) performing pixel-level integration of SOC stocks constrained by the predicted ST to avoid carbon assignment to non-soil layers. The ST model achieved a validation R2 of 0.57, revealing a distinct spatial pattern of deeper soils in the north and south and shallower soils in the central Plateau, with a mean depth of 90.40 ± 35.50 cm. The SOC model achieved a validation R2 of 0.56 for the topsoil (0–30 cm), with accuracy decreasing in deeper layers due to sparser observations and reduced influence of surface-based covariates. After ST-constrained integration, the total SOC stock within the reliable prediction area (area of applicability, AOA) was 4.72 Pg C (90 prediction interval: 0.99–19.43 Pg C), while the total stock across Tibet ranged from 2.58 to 47.50 Pg C for the 0–100 cm layer. These findings demonstrate that explicitly incorporating ST and ESM effectively eliminates the systematic biases inherent in the FD approach, offering a robust and scalable framework for accurate carbon accounting in complex, high-relief terrains.

Accurate estimation of gross primary productivity (GPP) is crucial for understanding terrestrial carbon cycles and assessing ecosystem health. Light use efficiency (LUE) models, which are widely used for the generation of regional or global GPP products, often rely on the fraction of absorbed photosynthetically active radiation (FAPAR). However, most of the existing FAPAR products with moderate to coarse spatial resolutions introduce uncertainties in GPP estimations across heterogeneous landscapes. In this work, the MODerate resolution Imaging Spectroradiometer (MODIS) FAPAR product at the 500-m resolution, along with a new HIgh-spatial-resolution Global LAnd Surface Satellite (Hi-GLASS) FAPAR dataset at the 30-m resolution, was used to drive an LUE model for GPP estimations at 188 eddy covariance (EC) sites. Then, they were compared and evaluated based on the EC GPP measurements. Results showed that Hi-GLASS FAPAR provided the GPP estimates with more detailed spatial information compared with MODIS FAPAR. Moreover, Hi-GLASS FAPAR significantly improved GPP estimations, with an overall R-2 increase from 0.54 (MODIS) to 0.63 (Hi-GLASS) and a root-mean-square error (RMSE) decrease from 3.04 to 2.70 gC & sdot;m(-2)& sdot;day(-1). In addition, 75% of the selected sites exhibited enhanced R-2 values with Hi-GLASS FAPAR, demonstrating its application potential in GPP estimations across different vegetation types. Specifically, crop sites exhibited the most notable improvements, with an R-2 increase of 0.16 and an RMSE decrease of 0.70 gC & sdot;m(-2)& sdot;day(-1). These findings highlight the advantages of high-resolution FAPAR data in capturing spatial heterogeneity and improving the accuracy of GPP estimations and underscore its potential for refined ecosystem monitoring.

Soil thickness (ST) is a critical indicator of ecological processes such as water regulation, nutrient cycling, carbon storage, and vegetation restoration. However, digital mapping of ST is often hindered by prediction bias and difficulties in identifying driving factors, largely due to the prevalence of right-censored data. Here, we investigate the eastern Himalayas (500-7000 m a.s.l.) using 130 field profiles and a suite of environmental variables spanning lithology, vegetation, climate, topography, spectral indices, and soil properties. We propose a composite framework that integrates the inverse probability of censoring weighted random forest (IPCW-RF) with Shapley additive explanations (SHAP) analysis. The IPCW component corrects censoring bias, the RF model provides high-precision spatial prediction, and SHAP yields quantitative insights into the mechanisms shaping ST at global, local, and spatial scales. The IPCW-RF model, trained with climate, topographic, and Sentinel-2 spectral variables, achieved strong predictive performance (R-2 = 0.55; MAE = 18.09 cm; RMSE = 21.59 cm) under 5-fold nearest-neighbor distance matching cross-validation (CV). Areas with high ST values were concentrated in the Yarlung Tsangpo River valley and on gently sloping plateau surfaces. The relationship between ST and elevation followed a nonlinear but structured pattern: ST decreased significantly between 500 and 2500 m, fluctuated between positive and negative correlations from 2500 to 5000 m, and declined again above 5000 m. SHAP analysis revealed an elevation-dependent contribution of environmental factors, with Band 8 emerging as the dominant predictor overall. At low to mid elevations (500-2500 m), vegetation played the primary role; at mid to high elevations (2500-4500 m), both vegetation and topography were influential; and at high elevations (>4500 m), topographic controls predominated. This study demonstrates the integration of an interpretable machine-learning framework with censored data, offering new insights into soil formation processes and improving spatial prediction of ST in complex plateau terrains.

Soil organic carbon (SOC) in the active layer (0–2 m) of the Tibetan Plateau (TP) permafrost region is sensitive to climate change, with significant implications for the global carbon cycle. Environmental factors—including parent material, climate, vegetation, topography, soil, and human activities—inevitably drive SOC variations. However, vegetation and climate are likely the two most influential factors impacting SOC variations. To test this hypothesis, we conducted experiments using 31 environmental variables combined with the recursive feature elimination (RFE) algorithm. These experiments showed that RFE retained all vegetation variables [Land cover types (LCT), normalized difference vegetation index (NDVI), leaf area index (LAI), and gross primary productivity (GPP)] as well as two climate variables [Moisture index (MI) and drought index (DI)], supporting our hypothesis. We then analyzed the relationship between SOC and the retained vegetation and climate variables using random forest (RF), Shapley additive explanations (SHAP), and GeoDetector models to quantify the independent and interactive drivers of SOC distribution and to identify the optimal conditions for SOC accumulation. The RF model explained 68

Solar farms have been rapidly expanding on the Qinghai-Tibetan Plateau.However,the effects of photovoltaic arrays on the contribution of microbial necromass carbon(MNC)to soil organic carbon(SOC),along with the underlying mechanisms,remain unclear.To address this,we collected soil samples from the top 20 cm in under-panel,inter-panel and control plots at five solar farms constructed between 2012 and 2014 in the dry Yarlung Tsangpo and Lhasa River valleys on the Qinghai-Tibetan Plateau.We determined SOC,fungal and bacterial necromass and relevant soil properties.We found that the concentration of MNC in the under-panel plots(3.93±0.79 mg g-1)was significantly higher compared to the control plots(2.28±0.79 mg g-1)across all five solar farms.The proportion of MNC to SOC in the under-panel plots(34.7±2.4％)was also significantly higher than that in the control plots(27.5±1.4％).Specifically,the contribution of fungal necromass to SOC in the under-panel plots(26.4±2.2％)was significantly larger than that in the control plots(19.7±1.6％),while the increase in the bacterial necromass proportion was insignificant.Partial least squares structural equation modeling(PLS-SEM)indicated a significant and positive effect of increased soil moisture in the under-panel plots on the proportion of fungal necromass to SOC.These results highlight that beyond their economic benefits,solar farms in the arid regions on the Qinghai-Tibetan Plateau can enhance soil C sequestration by improving soil moisture and promoting microbial necromass accumulation.

The Hengduan Mountains in Southwest China are currently facing a recurring and escalating threat of forest fires, posing a significant challenge to vegetation restoration efforts given the ecological value of this region. However, the effects of artificial restoration compared to natural regeneration on vegetation recovery dynamics have been less explored. This study focused on two sites in the Hengduan Mountains, primarily composed of conifer forests. One site employed a mix of artificial restoration measures, including tree planting, aerial grass seeding, and fertilizer application. The other site, which is relatively drier and colder, relied on natural regeneration. Four fire events in these two sites were compared to examine the effects of artificial restoration on vegetation recovery using remotely sensed data from Landsat and MODIS. Fire severity was classified using the differenced Normalized Burn Ratio (dNBR) and validated with field samples. Vegetation greenness loss and recovery trajectory were quantified using the near-infrared reflectance of vegetation (NIRv). The effects of artificial restoration were further analyzed after employing unburnt controls to minimize the impact of climate fluctuations. Results showed that fire severity mapping achieved an overall accuracy of 82%. The average NIRv loss across the four fires was 30%, 49%, and 69% in areas of low, moderate, and high severity, respectively. In low severity areas, natural and artificial recovery rates were similar, at 21% and 26% respectively, three years post-fire. In moderate severity areas, artificial restoration achieved a 46% recovery rate compared to 31% for natural recovery. In high severity areas, artificial restoration resulted in a 63% recovery rate versus 45% for natural recovery. Although the differences in recovery patterns may also be attributed to inherent variations in the vegetation composition, physical environment, and their interactions, this study suggests that high severity areas benefit significantly from artificial intervention for rapid recovery. Future studies should include more controlled experiments at the plot scale to further elucidate the processes of vegetation response to different artificial restoration measures.

Fire severity mapping can capture heterogeneous fire severity patterns over large spatial extents. Although numerous remote sensing approaches have been established, regional-scale fire severity mapping at fine spatial scales (<5 m) from high-resolution satellite images is challenging. The fire severity of a vast forest fire that occurred in Southwest China was mapped at 2 m spatial resolution by random forest models using Sentinel 2 and GF series remote sensing images. This study demonstrated that using the combination of Sentinel 2 and GF series satellite images showed some improvement (from 85% to 91%) in global classification accuracy compared to using only Sentinel 2 images. The classification accuracy of unburnt, moderate, and high severity classes was significantly higher (>85%) than the accuracy of low severity classes in both cases. Adding high-resolution GF series images to the training dataset reduced the probability of low severity being under-predicted and improved the accuracy of the low severity class from 54.55% to 72.73%. RdNBR was the most important feature, and the red edge bands of Sentinel 2 images had relatively high importance. Additional studies are needed to explore the sensitivity of different spatial scales satellite images for mapping fire severity at fine spatial scales across various ecosystems.

为阐明火烧强度对西南亚热带山地森林土壤理化性质的影响,本研究通过采集和分析四川西昌"3·30"火烧迹地不同火烧程度(未过火、低度、中度和重度)的土壤,发现重度火烧极显著增加了土壤容重;火烧显著增加了土壤速效磷和速效钾含量,显著降低了土壤全氮和碱解氮含量,但是对土壤pH值、有机碳、全钙、全镁、全钾和全磷含量无显著影响.通过对比分析火烧对不同纬度林地土壤理化性质的影响,发现火烧对亚热带地区森林土壤的影响与北方存在差异,主要表现为火烧后进入土壤的碱基离子和未完全燃烧的有机物相对较少,导致有机碳、全钙、全镁和全钾含量的变化不显著,而高温导致碱解氮含量显著降低.本文的结果不仅凸显了林火对不同气候条件下森林土壤性质影响的差异,而且可为亚热带地区火烧迹地植被恢复过程中的养分管理提供科学依据.

The empirical retrieval method based on vegetation indices (VIs) has been extensively utilized to estimate the photosynthetic vegetation fractional cover (FPV) and non-photosynthetic vegetation fractional cover (FNPV). These indices, however, saturate in high biomass environments and are easily influenced by external factors. Three red edge (RE) bands (i.e., RE1, RE2, RE3) are available on the Sentinel-2 satellite, providing new options for estimating FPV and FNPV. Here, sensitivity analysis from PROSAIL-PRO simulations provided a theoretical foundation for developing new indices. Sentinel-2 images and field observations were collected at three growth stages to test the original and new-developed indices for FPV and FNPV estimations. Compared to the original photosynthetic vegetation indices (PVIs) containing the near-infrared (NIR) and red bands, the optimal combinations for FPV estimation in August were RE3 and red bands, while the combination of RE2 and RE1 bands performed best in April. By introducing RE3, RE2, RE1, and red bands in a certain proportion, 4-band red edge PVIs had the strongest correlation with FPV. Although the optimal 2-band non-photosynthetic vegetation indices (NPVIs) in November were two shortwave infrared (SWIR) bands, the combinations of NIR and RE3 bands performed best for FNPV estimation in April. Combining SWIR1, SWIR2, NIR, and RE3 bands, three 4-band red edge NPVIs were put forward, had the highest accuracy for estimating FNPV. The most prominent achievement of the red edge indices is to suppress undervaluation when vegetation cover is high while mitigating overestimation at low vegetation cover levels. However, the optimum weighting parameters of the improved VIs differed depending on different growth stages. We discovered novel 4-band red edge indices are useful for estimating FPV and FNPV.

青藏高原作为中低纬度地区最大的高山冻土区,多年冻土和季节冻土广泛分布.高精度的地表冻融监测结果对研究该区域的水热交换、碳氮循环和土壤冻融侵蚀非常重要.本文基于4个青藏高原典型地区的土壤温湿度观测网数据,开展利用LightGBM算法和随机森林算法进行土壤冻融循环监测的研究.在构建土壤冻融监测模型的过程中,发现土壤湿度是影响冻融判别的一个关键因子.使用AMSR2亮温数据和ERA5-Land土壤湿度数据,基于两种机器学习算法判别地表冻融状态,将结果与传统冻融判别式算法进行对比分析.结果表明:相比冻融判别式算法,LightGBM算法在白天和夜间的总体判对率提高了 12.09％;14.45％,随机森林算法在白天和夜间的总体判对率提高了 13.23％和14.96％.近80％的错分样本分布在-4.0℃～4.0℃之间,说明2个机器学习算法能够识别出稳定的土壤冻结状态和融化状态.另外,LightGBM算法和随机森林算法得到的日冻融转换天数的平均RMSE降低了 112.82和117.00;冻结天数的平均RMSE降低了 47.87和53.96;融化天数的平均RMSE降低了37.10和39.80.同时,基于随机森林算法计算了2014年7月-2015年6月青藏高原冻结天数、融化天数、日冻融转换天数.得到的青藏高原冻结天数图,以中国冻土区划图为参考进行精度评价,总体分类精度为96.78％.

Freeze‒thaw (FT) erosion has gradually become more severe due to climate warming, and concerns about FT erosion in ecologically fragile areas (e.g., high-altitude and high-latitude areas) continues to grow. Tibet, located at the Third Pole of Earth, is also in a substantial part underlain with seasonally frozen soil and subject to FT erosion. Evaluating the sensitivity and influential factors of FT erosion in Tibet is warranted to manage the ecological environment and human production activities. In this study, we investigated the sensitivity and spatial distribution characteristics of FT erosion in Tibet based on advanced remote sensing and geographic information system (GIS) technologies. To further explore the influence of each factor on FT erosion, we analyzed the sensitivity of FT erosion under each factor condition. Our results showed that the annual temperature range is the most influential factor on FT erosion among temperature, precipitation, topography and vegetation. In addition, we introduced the coefficient of variation (CV) to represent the stability of temperature and then used CMIP5 simulation data to estimate the susceptibility of FT erosion in Tibet over the next 30 years. The CVs in central and western Tibet were higher than those in other areas and thus need more attention to FT erosion in central and western Tibet in the future.

Land surface temperature (LST) has been used in many applications as its strong relationships with land surface processes. However, the greatest limitation on the use of LST is the missing data caused by cloud contamination and weather conditions. Although several studies have proposed empirical statistical methods to reconstruct all-weather LST at a moderate resolution (~1 km), these models cannot reflect the complex nonlinear relationships between LST and surface characteristics. Besides, most models also failed to take account for cloud radiative effect (CRE) on the LST. In this study, we first used XGBoost method to describe complex relationships of LST with surface characteristics from clear-sky pixels, and applied the model to retrieve hypothetical clear-sky LST under cloudy sky. Secondly, cloud radiative effect (CRE) on the LST was calculated based on energy balance using the National Centers for Environmental Prediction (NCEP) reanalysis data. The models were applied to reconstruct the all-weather LST in 2019 over the Tibetan Plateau. The spatial patterns of reconstructed LST indicated that our model could produce completely spatial-seamless LST and depict the detailed information. The accuracy of the XGBoost LST was validated against the clear-sky MODIS LST (average R 2 = 0.91, MAPE = 0.76, RMSE = 2.4 K). The validation of CRE-EB LST with in-situ measurements was separated into three conditions: clear sky (RMSE = 2.94 K-4.45 K, R 2 = 0.90-0.97), unclear sky (RMSE = 3.51 K-4.23 K, R 2 = 0.86-0.93), and overall (RMSE = 3.40 K-4.25 K, R 2 = 0.88-0.93). Additionally, we compared CRE-EB LST with GLDAS LST, for AR site, RMSE lower by 8.52 K; for DSL site, RMSE lower by 9.89 K; for JYL+YK site, RMSE lower by 1.23 K. The results revealed that CRE-EB LST provided higher accuracy, while the GLDAS obviously underestimated the LST over the TP. This study demonstrated the utility of proposed models to reconstruct all-weather LST.

The China–Pakistan Economic Corridor (CPEC) is one of the flagship projects of the One Belt One Road Initiative, which faces threats from water shortage and mountain disasters in the high-elevation region, such as glacial lake outburst floods (GLOFs). An up-to-date high-quality glacial lake dataset with parameters such as lake area, volume, and type, which is fundamental to water resource and flood risk assessments and prediction of glacier–lake evolutions, is still largely absent for the entire CPEC. This study describes a glacial lake dataset for the CPEC using a threshold-based mapping method associated with rigorous visual inspection workflows. This dataset includes (1) multi-temporal inventories for 1990, 2000, and 2020 produced from 30 m resolution Landsat images and (2) a glacial lake inventory for the year 2020 at 10 m resolution produced from Sentinel-2 images. The results show that, in 2020, 2234 lakes were derived from the Landsat images, covering a total area of 86.31±14.98 km2 with a minimum mapping unit (MMU) of 5 pixels (4500 m2), whereas 7560 glacial lakes were derived from the Sentinel-2 images with a total area of 103.70±8.45 km2 with an MMU of 5 pixels (500 m2). The discrepancy shows that Sentinel-2 can detect a large quantity of smaller lakes compared to Landsat due to its finer spatial resolution. Glacial lake data in 2020 were validated by Google Earth-derived lake boundaries with a median (± standard deviation) difference of 7.66±4.96 % for the Landsat-derived product and 4.46±4.62 % for the Sentinel-derived product. The total number and area of glacial lakes from consistent 30 m resolution Landsat images remain relatively stable despite a slight increase from 1990 to 2020. A range of critical attributes has been generated in the dataset, including lake types and mapping uncertainty estimated by an improved equation of Hanshaw and Bookhagen (2014). This comprehensive glacial lake dataset has the potential to be widely applied in studies on water resource assessment, glacial lake-related hazards, and glacier–lake interactions and is freely available at https://doi.org/10.12380/Glaci.msdc.000001 (Lesi et al., 2022).

The dry valley is a unique geographic phenomenon in Southwest China with severe water erosion. However, little is known regarding its dominant controls and the discrepancies between dry valley subtypes, leading to the poor management of water erosion. To solve these problems, the revised universal soil loss equation (RUSLE) and Geodetector method were used in a dry temperate (DT), dry warm (DW), and dry hot (DH) valley. Results indicated that dry valleys suffer severe water erosion with a value of 64.78, 43.85, and 33.81 t·ha−1·yr−1. The Geodetector method is proven to be an efficient tool to quantify the dominant factor of water erosion. It was established that land use types (LUT) have the closest relationship with water erosion. The controls for water erosion could be better explained by multi-factor interactions analysis, particularly for the combination of slope and LUT in DW (q = 0.71) and DH (q = 0.66). Additionally, regions at high risk of water erosion were characterized by steep slope (>30°) and low vegetation coverage (<50%) in DT, while the opposite is shown in DH. These findings could provide insight for guiding soil erosion management and ecological restoration strategies that balance economic and environmental sustainability.

植被盖度是刻画陆地生态系统植被覆盖的重要生态参量.以当雄县Landsat-8OLI为数据源,从10种常用植被指数中筛选出适合反演高寒草地生长季/非生长季草地植被盖度的植被指数,引入像元三分法确定端元特征值,通过不同植被指数基于像元二分模型反演植被盖度的对比分析,确定适合生长季/非生长季植被盖度最优植被指数,根据反演结果分析了研究区草地生长季/非生长季植被盖度的时空变化特征.结果表明:1)由可见光-近红外波段构建的植被指数适用生长季植被盖度反演,由短波红外构建的植被指数适用于非生长季植被盖度反演.2)基于MSACRI的像元二分模型适合非生长季植被盖度反演,基于NDVI的像元二分模型则最适用于生长季植被盖度的反演.3)研究区草地植被盖度随海拔增加呈现先增加后减少的单峰变化格局,草地集中分布于海拔4300～5100 m处.生长季植被盖度主要集中于20％～80％,非生长季绝大部分的草地盖度小于40％.研究结果可为草地生态系统碳存储、植被生产力、土壤侵蚀、生态水文等研究提供参考依据.