The Three-Rivers Headwater Region (TRHR) is located on the Tibetan Plateau, within a transitional zone between seasonally frozen ground and continuous permafrost. Over 70 % of the region is predominantly covered by alpine grasslands, a vulnerable ecosystem increasingly threatened by ongoing permafrost degradation. This study utilized satellite data to analyze permafrost degradation by measuring active layer thickness (ALT) and the soil non-frozen period (NFP), and to investigate their impacts on alpine grassland growth. Results showed significant permafrost degradation from 2000 to 2020, with ALT thickening at a rate of 7.79 cm/decade (p < 0.05) and NFP lengthening by 1.1 days/yr (p < 0.05). Simultaneously, grassland vegetation exhibited a significant greening trend (0.0014 yr(-1), p < 0.01). Using the partial least squares (PLS) regression method, the study evaluated the relationships between grassland dynamics and permafrost degradation, while jointly accounting for climate variables (temperature, precipitation, and sunshine duration). ALT thickening was the dominant explanatory variable for grassland growth in 11.09 % of the region, and it was positively correlated in relatively cold western and alpine areas, but negatively correlated in the relatively warm eastern and central regions. NFP extension was the dominant explanatory variable for grassland growth in 10.38 % of the region, although its positive correlation weakened as climate conditions transitioned from relatively cold-dry to relatively warm-wet. Although permafrost degradation was positively correlated with grassland greening in relatively cold regions, the diminishing benefit of NFP extension and the adverse effects of ALT thickening may increasingly undermine grassland stability in relatively warm regions under further climate warming.

The Tibetan Railway has introduced pressures on the fragile grassland ecosystems of the Tibetan Plateau. However, the impact of the railway on the carbon sequestration remains unclear, as existing studies primarily focus on in-situ vegetation observations. In this study, we extracted the start and end of the growing season (SOS, EOS) and maximum daily GPP (GPPmax) along the railway corridor from the satellite-derived Gross Primary Productivity (GPP) data, and quantified the extent and intensity of the railway's disturbance on these indicators. We further employed the Statistical Model of Integrated Phenology and Physiology (SMIPP) to translate these disturbances into annual cumulative GPP (GPPann). Results show that Tibetan Railway significantly influences grassland within 50-meters, causing earlier SOS (0.1086 d m-1), delayed EOS (0.0646 d m-1), and reduced GPPmax (0.0069 gC m-2 d-1 m-1) as the distance to the railway gets closer. The advanced SOS and delayed EOS contributed gains of 28.82 and 104.26 MgC y-1, but reduction in GPPmax accounted for a loss of 2952.79 MgC y-1. Railway-induced phenology-physiology trade-off causes GPPann loss of 2819.71 MgC y-1. This study reveals Tibetan Railway's impact on grassland carbon cycling, offering insights for grassland conservation and sustainable transportation infrastructure projects.

Vegetation greening across the Tibetan Plateau, a critical ecological response to climate warming and land-cover change, affects soil hydrothermal regimes, altering soil moisture (SM) and soil temperature (ST) dynamics. However, its effects on SM-ST coupling remain poorly understood. Using integrated field measurements from a vegetation-soil (V-S) network, reanalysis, and physics-based simulations, we quantify responses of SM, ST, and their coupling to vegetation changes across the Upper Brahmaputra (UB) basin, southern Tibetan Plateau. Results show that strong positive SM-ST correlations occur throughout 0-289 cm soil layers across the basin, consistent with the monsoon-driven co-occurrence of rainy and warm seasons. Spatially, SM-ST coupling strength exhibits pronounced spatial heterogeneity, demonstrating strongest coupling in central basin areas with weaker intensities in eastern and western regions. Overall, vegetation greening consistently induces soil warming and drying: as leaf area index (LAI) increases from 20 % to 180 % of its natural levels, SM (0-160 cm) declines by 15 % to 29 % due to enhanced evapotranspiration and root water uptake. Mean ST simultaneously increases by 1.4 +/- 0.9 degrees C. Crucially, sparsely vegetated regions sustain warming (1.4-2.1 degrees C), while densely vegetated areas transition from initial warming to gradual cooling. These findings advance our understanding of soil hydrothermal dynamics and their broader environmental impacts, improving climate model parameterizations and informing sustainable land management strategies in high-altitude ecosystems.

The freeze-thaw erosion zone of the Tibetan Plateau (FTZTP) maintains an ecologically fragile system with enhanced thermal sensitivity under climate warming. Vegetation phenology in this cryosphere-dominated environment acts as a crucial biophysical indicator of climate variability, showing potentially amplified responses to environmental changes relative to other ecosystems. To investigate vegetation phenological characteristics and their climate responses, we derived key phenological parameters (the start, end and length of growing season-SOS, EOS, LOS) for the FTZTP from 2001 to 2021 using MODIS EVI data and analysed their spatiotemporal patterns and climatic drivers. Results indicated that the spatial distribution of phenology was highly heterogeneous, influenced by local climate, complex topography and diverse vegetation. SOS generally exhibited a delayed trend from east to west, while EOS was progressively later from the central plateau towards the southeast and southwest. Consequently, LOS shortened along both east-west and south-north gradients. Under sustained warming and wetting, the region experienced intensified freeze-thaw cycles, characterised by a delayed freeze-start, advanced thaw-end and shortened freeze-thaw duration. Both climate warming and freeze-thaw changes drove an overall significant advancement of SOS (-3.1 days/decade), delay of EOS (+2.2 days/decade) and extension of LOS (+5.3 days/decade) over the 21-year period. Notably, an abrupt phenological shift occurred around 2015. Prior to 2015, both SOS and EOS advanced, whereas afterward, SOS transitioned to a delaying trend and EOS exhibited a markedly stronger delay, leading to a pronounced extension of LOS. This regime shift was primarily attributed to changes in hydrothermal conditions controlled by climate warming and evolving freeze-thaw dynamics, with temperature being the dominant factor and precipitation exerting seasonally differential effects. Our findings elucidate the complex responses of alpine cryospheric ecosystems to climate change, revealing freeze-thaw processes as a key modulator of vegetation phenology.

Permafrost degradation under climate warming plays a crucial role in hydrological and ecological processes, including the regional water cycle and terrestrial carbon balance. The Tibetan Plateau (TP), which contains the largest expanse of high-altitude permafrost globally, remains understudied in terms of how permafrost degradation affects surface water resources and regional carbon dynamics. Using permafrost simulation models and quantitative analysis, we assess the spatiotemporal impacts of permafrost degradation on surface water resources and carbon dynamics. In the inner endorheic regions of the TP, ground ice meltwater contributed 12.6% of the total lake volume increase from 2000 to 2020, accelerating lake expansion and affecting nearby infrastructure and ecosystems. Cryospheric meltwater accounted for 4.6% of total runoff in the source areas of the Yangtze, Yellow, Lancang, Yarlung Zangbo, and Nujiang Rivers in 2002-2018. This cryospheric meltwater contribution is projected to peak in the 2030s-2040s, followed by a decline, with potentially profound implications for downstream water availability. From 2000 to 2020, carbon sequestration of alpine grassland in permafrost regions is 1.05-1.29 Tg C a-1 in 2000-2020. This estimate is underestimated by approximately 35.5% to 48.1% without considering the impact of permafrost degradation. Top-down thawing of permafrost from 2002 to 2050 is projected to release 129.39 +/- 21.02 Tg C a-1 of thawed soil organic carbon (SOC), with 20.82 +/- 3.06 Tg C a-1 decomposed annually. Additionally, permafrost collapse and thermokarst lake are estimated to reduce ecosystem carbon sinks by 0.41 (0.29-0.52) Tg C a-1 in 2020. (c) 2025 The Authors. Published by Elsevier B.V. and Science China Press. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Thawing permafrost alters climate not only through carbon emissions but also via energy-water feedback and atmospheric teleconnections. This review focuses on the Tibetan Plateau, where strong freeze-thaw cycles, intense radiation, and complex snow-vegetation interactions constitute non-carbon climate responses. We synthesize recent evidence that links freeze-thaw cycles, ground heat flux dynamics, and soil moisture hysteresis to latent heat feedback, monsoon modulation, and planetary wave anomalies. Across these pathways, both observational and simulation studies reveal consistent signals of feedback amplification and nonlinear threshold behavior. However, most Earth system models underrepresent these processes due to simplifications in freezethaw processes, snow-soil-vegetation coupling, and cross-seasonal memory effects. We conclude by identifying priority processes to better simulate multi-scale cryosphere-climate feedback, especially under continued climate warming in high-altitude regions.

Frozen soil, covering most of the Tibetan Plateau (TP), critically influences land surface and climate simulations. Although some studies have made advancements in simulations, further investigation into the distinct mechanisms underlying relevant parameterization schemes remains essential. This study compares two frozen soil permeability schemes in Noah-MP (NY06: high-permeability; Koren99: low-permeability) to elucidate their distinct hydrological mechanisms. Although significant disparities exist in the simulation of soil water and ice content between the two schemes in permafrost regions, the simulated soil water content in the shallow layer exhibits similarity. Their underlying physical processes behind this similarity differ fundamentally: Koren99 relies on cross-seasonal ice melt recharge, whereas NY06 depends more on current-season precipitation and snowmelt. With greater soil depth, soil water differences progressively propagate downward, amplifying variations in hydraulic conductivity, and soil memory effects become increasingly dominant. Meanwhile, the Koren99 scheme more effectively impedes bottom-up melting water transport than top-down effect. However, the aforementioned disparities are not apparent in seasonally frozen soil. Notable disparities also exist in simulated evapotranspiration and surface runoff over permafrost regions, particularly during the summer months. This research investigates the differences in water transport within frozen soil over the TP, elucidates the distinct hydrological mechanisms underlying different frozen soil permeability schemes, and highlights that similar soil hydrothermal simulations are associated with different physical processes, leading to varying degrees of effectiveness in soil memory. Furthermore, this research elucidates the dual role of soil ice (permeability restriction and water storage) in hydrological processes, providing a theoretical basis for improving frozen soil parameterization.

Permafrost degradation on the Tibetan Plateau (TP) has triggered widespread retrogressive thaw slumps (RTSs), affecting hydrology, carbon sequestration and infrastructure stability. To date, there is still a lack of long-term monitoring of RTSs across the TP, the thaw dynamics and comprehensive driving factors remain unclear. Here, using time-series Landsat imagery and change detection algorithm, we identified RTSs on permafrost regions of the TP from 1986 to 2020. Existing RTSs inventories and high-resolution historical imagery were employed to verify the identified results, the temporal validation of RTSs disturbance pixels demonstrated a high accuracy. In the study area, a total of 3537 RTSs were identified, covering a total area of 5997 ha, representing a 26-fold increase since 1986, and 69.2 % of RTSs formed since 2010. Most RTSs are located on gentle slope (4-12 degrees) at elevations between 4500 m and 5300 m, with a tendency to form in alpine grassland and alpine meadow. Annual variations in RTSs area exhibited a significant positive correlation with minimum air temperature, mean land surface temperature, and annual thawing index, while it showing a significant negative correlation with the decrease in downward shortwave radiation. Spatially, RTSs were more common in areas with higher soil water content and shallower active layer. Landsat imagery captured the vast majority of RTSs on the TP and revealed interannual disturbance details, but the 30 m resolution remains inadequate for delineating the refined boundaries of some micro-scale (< 0.18 ha) RTSs. Detected RTSs disturbances on the TP will aid in hazard management and carbon feedback assessments, and our findings provide novel insights into the impacts of climate change and permafrost environments on RTSs formation.

Study area: The Binggou and adjacent Yakou catchments in the northeastern Tibetan Plateau. Study focus: Hillslope flow paths were studied using hydrochemical data of various water types in the spring snowmelt and summer rainfall periods based on hydrochemical tracers and endmember mixing analysis. New hydrological insights for the study region: End-member mixing analysis confirmed the dominance of surface and near-surface runoff during the spring snowmelt. Specifically, the spring Binggou stream water had 61 % surface runoff, 22 % shallow groundwater, and 17 % near-surface runoff. The spring Yakou stream water had 64 % snowmelt, 25.5 % near-surface runoff, and 10.5 % riparian saturated soil water at a depth of 20 cm. The application of end-member mixing analysis failed in the summer rainfall period, and shallow subsurface flow contributed the most to the streamflow (similar to 100 %). The average acid-neutralizing capacity of the spring Yakou stream water was 611 mu eq/L, increasing to 841 mu eq/L in the summer, and for the Binggou stream water, the values were 747 mu eq/L and 1084 mu eq/L, respectively, indicating that the thawed soil layers had a significant buffering effect on stream water chemistry. This study revealed seasonal shifts in flow paths and stream sources, with a transition from surface to subsurface flow influenced by meteorological conditions and the active layer thickness. Future climate change may enhance subsurface flow recharge, leading to less diluted streamflow and stronger water-soil interactions.

Ice records provide a qualitative rather than a quantitative indication of the trend of climate change. Using the bulk aerodynamic method and degree day model, this study quantified ice mass loss attributable to sublimation/evaporation (S/E) and meltwater on the basis of integrated observations (1960-2006) of glacier-related and atmospheric variables in the northeastern Tibetan Plateau. During 1961-2005, the average annual mass loss in the ice core was 95.33 +/- 20.56 mm w.e. (minimum: 78.97 mm w.e. in 1967, maximum: 146.67 mm w.e. in 2001), while the average ratio of the revised annual ice accumulation was 21.2 +/- 7.7% (minimum: 11.0% in 1992, maximum 44.8% in 2000). A quantitative formula expressing the relationship between S/E and air temperature at the monthly scale was established, which could be extended to estimation of S/E changes of other glaciers in other regions. The elevation effect on alpine precipitation determined using revised ice accumulation and instrumental data was found remarkable. This work established a method for quantitative assessment of the temporal variation in ice core mass loss, and advanced the reconstruction of long-term precipitation at high elevations. Importantly, the formula established for reconstruction of S/E from temperature time series data could be used in other regions.