The Qinghai-Tibet Plateau (QTP) is a cornerstone of global climate and terrestrial ecosystems, playing a pivotal role in carbon cycling, water balance, and energy exchange. While research on QTP vegetation dynamics has traditionally focused on temperature (Ta) and precipitation (PRE), the independent effects of vapor pressure deficit (VPD) and soil moisture (SM) remain largely obscured, primarily due to strong hydrothermal coupling and pervasive multicollinearity among these variables. Therefore, we employed ridge regression to quantify the independent contributions of key hydrothermal factors to fractional vegetation cover (FVC) dynamics across the QTP from 1982 to 2020. The principal results are summarized as follows: (1) Spatiotemporally, FVC exhibited a distinct southeast-to-northwest declining gradient. Over the period 1982-2020, a consistent greening trend was observed at a rate of 0.0006 yr(-1). Specifically, vegetation greening occurred across 66.07% of the plateau, primarily in the arid and transitional zones, contrasting with localized vegetation degradation in the relatively humid zone. (2) Among the hydrothermal factors examined, VPD surpassed temperature to emerge as the dominant driver, explaining 31.69% of the FVC variation and controlling 44.31% of the total area. Ta followed, with a relative contribution of 24.15% and a dominant area of 27.40%. Deep soil moisture (SM3) accounted for 14.75% of the FVC trend, with a dominant area of 13.76%, exerting a critical influence in the arid zone. (3) FVC responses to VPD were highly heterogeneous across hydrothermal regimes: increasing soil moisture mitigated atmospheric drought stress to sustain slow greening in the arid zone; moderate increases in VPD promoted significant greening under adequate moisture in the transitional zone; whereas the synergistic interplay of high VPD and reduced soil moisture intensified vegetation decreasing in the relatively humid zone on the QTP. By disentangling the complex interplay between atmospheric and soil moisture, this study highlights the spatially heterogeneous sensitivity of alpine vegetation to hydroclimatic shifts, offering essential scientific evidence for ecosystem-specific conservation and adaptive management under global warming.

Most endorheic lakes on the Tibetan Plateau (TP) have expanded rapidly since the late 1990s, causing significant impact on the regional eco-environment. While changes in lake storage have been extensively investigated, the limnological processes in response to this rapid growth remain unclear. Based on eight times of in situ observations between 2009 and 2025, we analysed changes in stratification of Dagze Co, a meromictic lake on the central TP. Over the past 16 years, the lake level increased cumulatively by 6.5 m, accompanied by a decrease in salinity in both the epilimnion (from 14.8 to 12.3 psu) and hypolimnion (from 19.4 to 17.3 psu). Although the epilimnion thickness fluctuated significantly, its absolute position shifted upwards with the rising water level. The hypolimnion was less affected by lake level increase due to the persistent density gradient. Consequently, the metalimnion thickened considerably, characterized by the expansion of both thermocline and chemocline. A secondary density stratification emerged following an abrupt lake-level increase in 2017-2018, which more than doubled the metalimnion thickness. The secondary stratification weakened but persisted in subsequent years, with the upper boundary of chemocline descending by similar to 7.0 m from 2018 to 2023. Our results indicate that climate change and the rapid lake expansion have significant impact on the stratification regime of lakes on the TP. More attention should be paid to understanding the responses of lake ecosystems to changes in salinity and stratification.

Fire activity plays a crucial role in carbon emissions and climate change worldwide. Fragile ecosystems in arid Central Asia (ACA) are sensitively affected by fire activity. However, little is known about the patterns and driving forces of past grassland fires in ACA due to the scarcity of sedimentary records. Here, we analyzed concentration and influx of charcoal from the sediment of Lake Sasikul (Tajikistan), to reconstruct the regional and local fire activities on the Pamir Plateau over the past 2500 years. The regional fire activity reconstructed by micro-charcoals (20-50 μm) presents a long-term increasing trend. Comparisons with paleo-vegetation, paleo-climate, and historical data suggest that the regional fire activity before 1000 cal yr BP mainly responded to climate change, while it was more controlled by human activities after 1000 cal yr BP. In addition, the contribution of vegetation (fuel) conditions to regional fire activity has been relatively weak, except for the past few hundred years. Intriguingly, the local fire activity reconstructed from macro-charcoals (>125 μm) has a different trend from the regional one, with the highest level from 1750 to 600 cal yr BP. It is found that the local fire history could be mainly attributed to natural climate changes throughout the study period, consistent with the fact that there is little human impact near the lake even in the modern times. With current global warming and increasing human activities, fire prevention at various spatio-temporal scales faces greater pressure in arid regions. Therefore, clarifying the relationship between the fire activity, vegetation, climate, and human activities can provide a solid scientific basis for developing future wildfire management policies.

The Tibetan Plateau is a high-altitude arid region, where limited in-situ precipitation measurements are available. In this communication, we document a strong precipitation gradient at the southern edge of the Paiku Co catchment (southern Tibetan Plateau) from in-situ data and atmospheric model outputs. In particular, we use water pressure time series from proglacial lakes, two automatic weather stations, and data from ERA5-Land reanalysis and CORDEX-FPS-CPTP ensemble. We show that precipitation totals can vary by one order of magnitude over a short distance of 10 km in a rather smooth terrain during the cold season. This large precipitation gradient marks the transition between the great Himalayas and the Tibetan Plateau.

Previous studies reveal robust lake expansion on TP since the late 1990s. Debate on the quantitative contribution from precipitation and other factors including glacier melting continues, but with more evidence showing that precipitation dominates the interdecadal increase in inner TP lake water storage. However, the TP lake level changes (LLCs) and the potential connection with regional precipitation at interannual timescale and the underlying atmospheric processes, especially for the extreme years, remain unclear yet. The year 2015 and 2018 witness the largest lake level decrease and increase for some major TP lakes during past decades, accompanying extreme rainfall deficit and surplus in mid-summer (July-August) in these two years, respectively. Understanding such extreme hydroclimate events over TP is of great importance for the water resource security in Asia. The central to eastern TP (CETP) is divided into 8 subregions according to the geographical locations of data available lakes and meteorological stations. The mid-summer LLCs on central TP (CTP, subregion II) are robustly and highly related with the concurrent rainfall there, while LLCs on northern central TP (subregion III) are significantly linked to the rainfall over northeastern TP (NETP, subregion VI). The extreme TP lake level decrease and corresponding rainfall deficit in 2015 is closely related to the strong anomalous water vapor divergence, which is mainly induced by weakened westerly winds. Meanwhile, strong wave propagates from the Ural Mountains and North Atlantic to TP and East Asia. In contrast, the extremely high lake level increase and rainfall surplus in 2018 is predominated by easterly winds which are related to the strengthened East Asian summer monsoon. Simultaneously, the western Pacific subtropical high becomes much stronger and extremely northwestward. The CTP rainfall deficit year composite mostly reflects the 2015 case, and NETP rainfall surplus year composite largely resembles the 2018 case, showing similar but weaker atmospheric circulation signals.

The Tibetan Plateau(TP)and its surroundings are also referred to as"Asian Water Tower",which is the headwater of ten major rivers in Asia and provides ecosystem service for nearly two billion people[1].The interior TP boasts the greatest concentration of high-latitude inland lakes in the world,providing essential water resources,regulating regional climate and maintaining ecological balance.Over the past 50 years,the TP has experienced rapid cli-mate warming and overall wetting,which has led to glacier mass loss,thawing of permafrost,reduction in snow cover,and shorten-ing of lake ice duration[1].Changes in lake water storage provide a means of assessing the regional hydrological response to recent cli-mate change in this sparsely populated region.

Zige Tangco is a meromictic saline lake located on the central Tibetan Plateau. Two parallel cores (ZGTC A-1 and ZGTC A-2) were collected from the lake at a water depth of 25 m during summer 2006. The chronology of core A-1 was reconstructed based on the Constant Initial Concentration (CIC) model of 210 Pb and three accelerator mass spectrometry (AMS) ages from the chitin fragments. The hard water effect calibration of the sediment 14 C age showed that the reservoir effect ranged from 1655 yr at 1950 AD to 1540 yr at 1610 AD. The hydrological variation in Zige Tangco during the past 800 yr was reconstructed using multi-proxies, including organic and carbonate content, stable isotopes of fine-grained carbonate minerals (< 38.5 µm) and grain-size distribution of the lake sediments. Our results show that there were strong fluctuations in the lake level between 1200 and 1820 AD, and at least three dry periods were recorded between 1235 and 1315 AD, 1410 and 1580 AD, and 1660 and 1720 AD characterized by high carbonate content, abrupt positive shifts of stable isotopes, and high sand content. The low-lake-level periods during the Little Ice Age (LIA) in Zige Tangco correspond to the lower δ 18 O values in the Guliya ice core and the lower precipitation reconstructed from tree rings in Delingha. This demonstrated that the summer monsoon on the central Tibetan Plateau weakened during the dry and cold periods, whereas the winter monsoon strengthened. Relatively wetter periods or higher lake levels in Zige Tangco occurred at 1580–1650 AD and 1820–1900 AD. Negative shifts in stable isotopes were related to increased lake levels between 1800 and 1820 AD. Our results also showed that the summer monsoon precipitation on the central Tibetan Plateau was mainly controlled by solar activity during the past 800 yr.

The Asian drylands is a critical component of the Silk Road, comprising primarily the arid central Asia (ACA) and the arid west Asia (AWA). This region is highly sensitive to global changes, and the spatial patterns of climate change and variations in fire activity are prominent scientific concerns. We reconstructed the history of vegetation succession and climate change in the Asian drylands over the past 2000 years, explored the spatial differences in climate change, and revealed the history of fire activity and its influencing factors: In ACA, both LakeSasikul and Lake Bazangan records exhibit climatic characteristics of the Medieval Warm Period (MWP) aridity and a wetter climate during the Little Ice Age (LIA). However, in AWA, Almalou peat and its surrounding areas show climatic characteristics of a wetter MWP and a drier climate during the LIA. (2) On the centennial scale, in ACA and AWA have a dipole pattern of precipitation variations, which is consistent with the intergenerational precipitation changes in the global warming, revealing that this may be the fundamental feature of climate change in the Asian drylands. (3) The evolution of fire activity in the western section of the Silk Road has been gradually increasing over the past 2000 years. In Lake Bazangan, fuel (biomass) conditions are the main limiting factor for the evolution of fire activity in the natural state; in Lake Sasikul, the natural changes in climate dryness/wetness are the main controlling factors affecting the fire activity; in Amalou peat, the fire activity is affected by the fuel (biomass) and climate, showing a more complex response model.

Extreme events on the Tibetan Plateau (TP) have become increasingly frequent and severe in recent decades in tandem with rapid climate warming and moistening. Thus, it is highly important to investigate their processes, causes and environmental impacts. Using in-situ observations of lake level changes in combination with satellite altimetry data, we showed that lakes on the central TP (CTP) expanded unprecedentedly in 2017-2018, with lake level increases of 1.4-2.8 m at the five observed lakes. This extreme lake expansion occurred over almost the entire inner TP, which was mainly attributed to anomalously high precipitation (approximately 43-97% higher in 2017-2018 than in normal years on the CTP). The rapid lake expansion posed a substantial threat to regional ecosystems, infrastructure, and even residential habitats. Therefore, it is necessary to identify the potential sites threatened by lake expansion on the TP using high-resolution satellite images and perform near real-time ob-servations of lake level changes through satellite and in-situ observations.

Climate change modifies the water and energy fluxes between the atmosphere and the surface in mountainous regions such as the Qinghai–Tibet Plateau (QTP), which has shown substantial hydrological changes over the last decades, including rapid lake level variations. The ground across the QTP hosts either permafrost or is seasonally frozen, and, in this environment, the ground thermal regime influences liquid water availability, evaporation and runoff. Consequently, climate-induced changes in the ground thermal regime may contribute to variations in lake levels, but the validity of this hypothesis has yet to be established. This study focuses on the cryo-hydrology of the catchment of Lake Paiku (southern Tibet) for the 1980–2019 period. We process ERA5 data with downscaling and clustering tools (TopoSCALE, TopoSUB) to account for the spatial variability of the climate in our forcing data (Fiddes and Gruber, 2012, 2014). We use a distributed setup of the CryoGrid community model (version 1.0) to quantify thermo-hydrological changes in the ground during this period. Forcing data and simulation outputs are validated with data from a weather station, surface temperature loggers and observations of lake level variations. Our lake budget reconstruction shows that the main water input to the lake is direct precipitation (310 mm yr−1), followed by glacier runoff (280 mm yr−1) and land runoff (180 mm yr−1). However, altogether these components do not offset evaporation (860 mm yr−1). Our results show that both seasonal frozen ground and permafrost have warmed (0.17 ∘C per decade 2 m deep), increasing the availability of liquid water in the ground and the duration of seasonal thaw. Correlations with annual values suggest that both phenomena promote evaporation and runoff. Yet, ground warming drives a strong increase in subsurface runoff so that the runoff/(evaporation + runoff) ratio increases over time. This increase likely contributed to stabilizing the lake level decrease after 2010. Summer evaporation is an important energy sink, and we find active-layer deepening only where evaporation is limited. The presence of permafrost is found to promote evaporation at the expense of runoff, consistently with recent studies suggesting that a shallow active layer maintains higher water contents close to the surface. However, this relationship seems to be climate dependent, and we show that a colder and wetter climate produces the opposite effect. Although the present study was performed at the catchment scale, we suggest that this ambivalent influence of permafrost may help to understand the contrasting lake level variations observed between the south and north of the QTP, opening new perspectives for future investigations.

The ecological environment of arid central Asia (ACA) is fragile and sensitive to long-term climate change. Recent palaeoclimatological studies have mainly focused on northwestern China, which is located on the eastern side of the region. Holocene palaeoclimate records from the western region of ACA are scarce, thus hindering the exploration of the relationship between climate change and Silk Road civilization. In this study, we conduct a pollen analysis of Lake Sasikul on the Pamir Plateau in Tajikistan and use pollen data to quantitatively reconstruct the precipitation history over the last 2500 years. The results show that herbaceous pollen is primarily represented by Artemisia and Amaranthaceae, thus suggesting the persistent dominance of grassland in the vicinity of Lake Sasikul. Amaranthaceae, which is the most drought-tolerant pollen type, shows the highest values during the Medieval Warm Period (MWP, 950-1300 CE). The values of Artemisia and cold-wet-adapted Picea pollen are higher during the Little Ice Age (LIA, 1550-1900 CE). The quantitative reconstruction shows that during the MWP, the mean annual precipitation is 120 mm, which is approximately 15% lower than the level of modern precipitation; meanwhile, during the LIA, the average annual precipitation is 160 mm (up to 210 mm), which is approximately 20% higher than the present value. We combine our results with regional temperature records and archaeological data to discuss the possible effects of climate change on the development of Silk Road civilization. The hydrothermal configuration may have altered water resources and thus affect human activities in ACA. From 580 to 900 CE, i.e. during the Sui and Tang Dynasties, ice and snow meltwater increased under warm climate, whereas the amount of precipitation was average. Additionally, human settlements intensified along with urbanization, and the Silk Road civilization was prosperous and well-developed. From 1270 to 1650 CE, i.e. during the Yuan and Ming Dynasties, under overall colder and drier conditions and due to insufficient freshwater input, the intensity of local human settlement weakened, and the Silk Road civilization declined. Therefore, owing to global warming and increasing precipitation, new development opportunities have emerged for the development of agriculture and social economy in ACA.

After the considerable lake level decrease in 2015 in response to the super 2015/2016 El Niño and lake level recovery in 2016, an extreme lake expansion occurred on the central and northern TP in the following two years (2017 and 2018), in contrast with the slight lake level changes on the southern TP. In-situ observations at Zhari Namco near Cuoqing County show that lake level increased abruptly by 1.4 m and 1.7 m in summer 2017 and 2018, respectively, which was even close to the accumulated lake level increase between 2000 and 2015. At Dazeg Co near Nima Country, lake level increased by 0.9 m and 1.4 m in summer 2017 and 2018, respectively, which is about 3 times as large as the increasing rate between 2000 and 2015. At Eya Co and Cedo Caka near Shuanghu County, lake level accumulatively increased by 1.5 m and 2.0 m, respectively, in the two years. The extreme lake expansion had significant impact on geomorphology and even posed great threat on the infrastructures such as road and bridges around the lakes. Causes of the extreme lake expansion are investigated by examining precipitation data and changes in large scale circulations.

It is difficult to quantify the amount of groundwater inflow on the Tibetan Plateau (TP), yet it can be critically important for sustaining lake water balance. Here we show that most endorheic lakes on the western TP exhibited considerable water level increase during the ice‐covered period, which contrasts with lakes in other regions of the TP. An analysis of lake water balance attributes this water surplus to significant groundwater inflow, which is estimated to be about 59%–66% of total inflow into lakes. The groundwater inflow occurred after the 2000s, which is consistent with the rapid lake expansion and significant increase in precipitation. We suggest that the groundwater inflow is mainly related to large‐scale active faults in the limestone bedrock and sufficient meltwater from high elevations. Our results imply that groundwater may be deeply involved in the water cycle and modify the seasonal and inter‐annual lake variations on the western TP.

Ground thermal regime of high mountain catchments impacts the partition between infiltration and runoff, latent and sensible heat fluxes, frozen and liquid subsurface water and the presence (or absence) of permafrost. In the context of global warming, hydrological modifications associated to ground thermal changes are of critical importance for extensive headwater regions such as the Qinghai-Tibet Plateau (QTP) and the Himalayas, which are major water towers of the world. Improving our ability to quantify these changes is therefore a key scientific challenge both regarding basic science and continental-scale water resource management. Many watersheds of the QTP have seen their hydrologic budget modified over the last decades as evidenced by strong lake level variations observed in endorheic basins. Yet, the role of ground thermal changes in these variations has not been assessed. Lake Paiku (central Himalayas, southern TP) has exhibited important level decreases since the 70s and thus offers the possibility to test the potential role of ground thermal changes and permafrost thaw on these hydrologic changes. We present distributed ground thermo-hydric simulations covering the watershed over the last four decades to discuss their implications on the lake level changes. We use the Cryogrid model to simulate the surface energy balance, snow pack dynamics and the ground thermo-hydric regime while accounting for the phase changes and the soil water budget. Because the surface radiative, sensible and latent heat fluxes in alpine environments are strongly dependent on the physiography, the model is forced with distributed downscaled forcing data produced with the TOPOSCALE model to account for this spatial variability. Simulated surface conditions are evaluated against meteorological data acquired within the basin, ground surface temperature loggers and remotely sensed surface temperatures. The simulations show that, contrary to large scale estimates of permafrost occurrence probability, an significant part of the basin is underlaid by permafrost (>20%). We also show that over the 1980-2020 period, ground temperature warmed up by 1.5 to 2°C per centuries. The permafrost limit rose from 5100 to 5300 m asl (in 40 years). Unfrozen surface conditions increased by around 25 days per century and evaporation increasing by +22% over the period. To represent the impact of these changes on the lake level, we included them in a simple hydrological budget calculation including the contribution of glacier melt and lake evaporation. This approach shows that ground thermo-hydric changes in the catchment have significantly contributed to the lake level changes. These first results highlight the potential of thermo-hydric simulation to better quantify hydrological changes to come in the QTP.

Third Pole natural cascade alpine lakes (NCALs) are exceptionally sensitive to climate change, yet the underlying cryosphere-hydrological processes and associated societal impacts are largely unknown. Here, with a state-of-the-art cryosphere-hydrology-lake-dam model, we quantified the notable high-mountain Hoh-Xil NCALs basin (including Lakes Zonag, Kusai, Hedin Noel, and Yanhu, from upstream to downstream) formed by the Lake Zonag outburst in September 2011. We demonstrate that long-term increased precipitation and accelerated ice and snow melting as well as short-term heavy precipitation and earthquake events were responsible for the Lake Zonag outburst; while the permafrost degradation only had a marginal impact on the lake inflows but was crucial to lakeshore stability. The quadrupling of the Lake Yanhu area since 2012 was due to the tripling of inflows (from 0.25 to 0.76 km3/year for 1999 to 2010 and 2012 to 2018, respectively). Prediction of the NCALs changes suggests a high risk of the downstream Qinghai-Tibet Railway, necessitating timely adaptions/mitigations.

Lake surface water temperature (LSWT) is a key parameter in understanding the variability of lake thermal conditions and evaporation. The MODIS-derived LSWT is widely used as a reference for lake model validations and process studies in data-scarce regions. In this study, the accuracy of the MODIS LSWT was examined on the Tibetan Plateau (TP). In-situ subsurface temperatures were collected at five large lakes. Although the observation period covers from summer to winter, only the observations during the lake turnover period (from October to freeze-up), when the lakes are well mixed, can be used as ground truth. The MODIS LSWT agrees well with the selected in-situ data for the five large lakes, with root mean square error (RMSE) < 1 °C at nighttime and <2 °C in the daytime, indicating a high accuracy of the MODIS LSWT data. Before the turnover period, the water is thermally stratified and the surface water is warmer than the subsurface water, and thus the in-situ subsurface water temperature data and the MODIS LSWT have different representativeness. In this case, if the observations are used as a validation basis, the MODIS errors could be much magnified. This in turn indicates the importance of period selection for the validation.

The lower parts of two glaciers in the Aru range on the western Tibetan Plateau (TP) collapsed on 17 July and 21 September 2016, respectively, causing fatal damage to local people and their livestock. The giant ice avalanches, with a total volume of 150 × 106 m3, had almost melted by September 2019 (about 30 % of the second ice avalanche remained). The impact of these extreme disasters on downstream lakes has not been investigated yet. Based on in situ observation, bathymetry survey and satellite data, we explore the impact of the ice avalanches on the two downstream lakes (i.e., Aru Co and Memar Co) in terms of lake morphology, water level and water temperature in the subsequent 4 years (2016–2019). After the first glacier collapse, the ice avalanche slid into Aru Co along with a large amount of debris, which generated great impact waves in Aru Co and significantly modified the lake's shoreline and underwater topography. An ice volume of at least 7.1 × 106 m3 was discharged into Aru Co, spread over the lake surface and considerably lowered its surface temperature by 2–4 ∘C in the first 2 weeks after the first glacier collapse. Due to the large amount of meltwater input, Memar Co exhibited more rapid expansion after the glacier collapses (2016–2019) than before (2003–2014), in particular during the warm season. The melting of ice avalanches was found to contribute to about 23 % of the increase in lake storage between 2016 and 2019. Our results indicate that the Aru glacier collapses had both short-term and long-term impacts on the downstream lakes and provide a baseline in understanding the future lake response to glacier melting on the TP under a warming climate.

Evaporation from hydrologically closed lakes is one of the largest components of the lake water budget; however, its effects on seasonal lake-level variations remain unclear on the Tibetan Plateau (TP) due to a lack of comprehensive observations. In this study, weekly lake evaporation and its effects on seasonal lake-level variations are investigated at Paiku Co on the southern TP using in situ observations of thermal structure and hydrometeorology (2015–2018). Lake evaporation from Paiku Co was estimated to be 975±142 mm during the ice-free period (May to December), characterized by low values of 1.7 ± 0.6 mm d−1 during the pre-monsoon season (May to June), high values of 5.5±0.6 mm d−1 during the post-monsoon season (October to December), and intermediate values of 4.0±0.6 mm d−1 during the monsoon season (July to September). There was a ∼ 5-month lag between the maximum net radiation (June) and maximum lake evaporation (November). These results indicate that the seasonal pattern of lake evaporation from Paiku Co was significantly affected by the large lake heat storage. Contrasting hydrological and thermal intensities may play an important role in the large amplitude of seasonal lake-level variations at deep lakes like Paiku Co. High inflow from monsoon precipitation and glacier melting and moderate lake evaporation, for instance, drove rapid lake-level increase during the monsoon season. In contrast, high lake evaporation and reduced inflow caused lake level to decrease significantly during the post-monsoon season. This study implies that lake evaporation may play an important role in the different amplitudes of seasonal lake-level variations on the TP.

Region-wide investigation of glacier change in High Mountain Asia marked the strongest recession in southeastern Tibetan Plateau (SETP) in recent decades. However, evident differences of quantitative glacier mass balance (MB) estimations on fine scales exist in the prior reports. The large uncertainties in current geodetic observations over this region highlight the need for more independent validations and investigations on the spatial variability of glacier change. This study enriched glacier MB estimates over the SETP by using the newly released global Digital Elevation Model (DEM) - the TanDEM-X and analyzed the spatial pattern of glacier change on multi-spatial scales by comparing available satellite-based geodetic observations. Results reveal that the TanDEM-X - SRTM elevation differences, which show no obvious horizontal shifts and no significant elevation-dependent bias, provide reliable elevation change information over this challenging area. On the regional scale, the geodetic datasets (the ASTER elevation change rates data, TanDEM-X- SRTM differences, ICESat) achieved a highly consistent estimate of glacier MB at -4.11 +/- 1.52 Gt a(-1) (-0.60 +/- 0.16 m w.e. a(-1)) during 2000 and 2010s. On the subregional scale, the two gridded geodetic datasets revealed a clear contrast between moderate thinning of spring-accumulation type glaciers in the south and significantly more negative glacier MB of summer-accumulation type glaciers in the north. Glacier-wide MB showed high heterogeneity, which was closely related to the combined effect of topographic variation and climate forcings. We observed similar glacier thinning rates between debris-covered land-terminating glaciers and clean-ice glaciers (-0.50 +/- 0.32 m w.e. a(-1)) but remarkable more negative MB (-0.89 +/- 0.36 m w.e. a(-1)) of debris-covered glaciers connected with glacial lakes, which may indicate different evolution process of debris-covered glaciers in the monsoon-influenced region. This study demonstrates that current satellite-based geodetic observations achieve overall consistent patterns of glacier mass changes even over the most challenging mountainous areas (complicated terrain and climate). With the global TanDEM-X DEM and increasing observations from the mission, we expect improved consistency in glacier MB estimation on large spatial scales as well as detailed investigation of glacier mass change variability in the future.