Changes in glacier meltwater runoff in the Qilian Mountains situated on the northeastern edge of the Tibetan Plateau are important for sustaining water resources in the arid regions of Northwest China. Combining multi-period remote sensing, historical data and modeling, we evaluate the glacier mass and area changes as well as associated runoff change of all glaciers in the area. Glacier area shrunk by 516.8 km2 (similar to 26%, 0.53% +/- 0.15% a-1) between 1970 and 2020 and glacier surface elevation change was -0.35 +/- 0.04 m a-1 between 2000 and 2014. In general, glaciers in the eastern Qilian Mountains retreated faster than those in the western part. A mass balance model calibrated against geodetic estimates derived from DEM differencing was used to reconstruct annual glacier mass changes and runoff from 1990 to 2017. Across the 11 glacierized large-scale river basins (0.02%-2.5% glacierization), the index-based glacier runoff ratio relative to basin water input ranged from 0.1% to 6.2%. Results show an increase in glacier runoff volumes during 1990-2001, but no statistically significant trend thereafter (2002-2017) despite increasing glacier thinning. This slowdown is attributed to decreasing glacier area as the glaciers retreated. Overall, our findings highlight that while glacier runoff continues to play a critical role in sustaining river runoff in the Qilian Mountains, its buffering effect is weakening as glacier runoff approach peak water, implying increasing vulnerability of regional water resources to precipitation variability under future climate change.

Study region The Vakhsh River Basin, the largest tributary of the Amu River, is located in the Pamir Mountains in Central Asia. Research focus The Vakhsh River Basin, contributing ∼24% of the Amu Darya’s discharge and over 90% of Tajikistan’s hydropower, provides a key case for studying climate-driven streamflow changes. However, due to observational constraints, the coupling among glaciers, snow, and runoff in this region remains poorly understood. Therefore, this study employs the SWAT-Glacier-SRD model under a multi-objective calibration framework to investigate hydrological responses to climate change. New hydrological insights Results indicate that the model could well simulate hydrological processes in this catchment with daily NSE being 0.81 and 0.78 during the calibration and validation periods. During 1971–2019, streamflow increased significantly (1.57 mm yr⁻¹), while its composition remained stable, dominated by snowmelt (58.2%), followed by rainfall (30.5%) and glacier melt (11.3%). At the seasonal scale, the contribution of rainfall increased in early spring. Further analysis showed that streamflow timing is highly sensitive to climate conditions, with warmer, drier conditions leading to earlier and more dispersed flow, and colder, wetter conditions producing later and more concentrated flow. Despite significant warming in this region, streamflow timing shows only a slight, insignificant advance (−0.129 d yr⁻¹), likely due to compensatory precipitation effects, but this stability may not persist under intensified warming. This study provides a reliable framework for simulating cryosphere–hydrology processes in Central Asian basins.

Glacier collapse hazard events have become increasingly frequent on the Tibetan Plateau under climate warming. However, the spatiotemporal evolution of multi-source remote sensing-based precursory signals before glacier collapse remains poorly understood. In this study, multi-source remote sensing datasets acquired during 1990–2016 were used to systematically investigate the pre-collapse evolution of Aru 53 and Aru 50 Glacier in terms of glacier geometry, surface elevation, glacier surface velocity (GSV), surface albedo, glacier surface temperature (GST), and surface synthetic aperture radar (SAR) backscatter coefficient. The results show that both glaciers experienced continuous terminus retreat, accompanied by upstream surface lowering and downstream surface thickening within the collapse zone. GSV increased markedly. Prior to collapse, extensive crevasses developed in the central and frontal parts of Aru 53, accompanied by a progressive increase in SAR backscatter coefficient. The enhanced SAR signals spatially coincided with regions experiencing pronounced surface structural changes. In contrast, decreases in surface albedo and increases in GST did not exhibit distinctive pre-collapse signatures for either glacier. Overall, the characteristic pattern of upstream thinning coupled with downstream thickening, anomalous acceleration of GSV, and rapid crevasse development are identified as the key remote sensing precursors of glacier instability.

Coastal land and marine constraints mean that China can no longer rely solely on coastal nuclear power plants to meet inland load growth and regional balancing needs. Planned inland nuclear projects initiated in 2006 were suspended after Fukushima, heightening public expectations for nuclear safety and emergency response. Rapid and accurate estimation of source term release rates during inland nuclear accidents is therefore a key scientific issue for emergency decision making. This study uses the Pengze nuclear power plant in Jiangxi Province as a representative inland site and establishes a 30 km & times; 40 km modeling domain centered on the plant. Three typical accident scenarios are constructed, and three corresponding time-series source term inversion models are developed. Using I-131 as a tracer, ERA5 data are processed with CALMET to build realistic meteorological fields, and CALPUFF is used to generate 5-h, minute-resolution concentration data. These data train a gated recurrent unit (GRU) model with attention pooling to achieve minute-by-minute inversion of I-131 release rates, and its performance is systematically compared with that of baseline GRU and long short-term memory (LSTM) sequence models under the three scenarios. Sensitivity analysis and explainable analysis techniques are further applied to enhance physical transparency and increase confidence in using the inversion results to support nuclear emergency management.

Why rivers have not deeply incised the interior of the Tibetan Plateau over geologic time scales remains an open question. Resolving this issue requires clarifying and quantifying the respective influences of different geological factors on headward river erosion. Focusing on the Eastern Himalayan syntaxis, we reveal that glacial damming occurred frequently along an ∼70 km reach of the Yarlung Tsangpo River, based on 10Be exposure dating and paleoglacial modeling. Glacial-damming episodes limited headward erosion via a cycle in which prolonged damming inhibited erosion during glacial periods and short-lived damming delayed it during interglacial periods. Moreover, these damming processes triggered extensive aggradation upstream and protected bedrock from erosion over long time scales, while they enhanced downstream incision, reshaping the longitudinal profile of the Yarlung Tsangpo River. Repeated glacial-damming processes cumulatively raised the effective local base level by ∼300 m since the middle Pleistocene. Overall, we show that the coupling of glaciation and tectonics is the key factor maintaining the long-term stability of the southeastern Tibetan Plateau margin.

This study proposes a source term inversion method for nuclear accidents based on the Harris Hawks Optimization (HHO) algorithm and a Gaussian plume model, enabling accurate estimation of radionuclide release rates and the two-dimensional location of release points using off-site monitoring data under accident scenarios. To evaluate model performance, validation was conducted through simulated experiments under two accident scenarios with known and unknown release locations and tracer experiments involving seven different release scenarios. The simulation results demonstrate that, compared with two other swarm intelligence algorithms, Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), the HHO-based inversion model achieves higher estimation accuracy, faster convergence speed, and greater stability during iterative inversion. The convergence rate and accuracy of the model are somewhat dependent on the initialization range of the population and the boundary constraints of the target parameters. The tracer experiment validation shows that the HHO model performs well in most cases, with an average relative error of 0.0341 in release rate inversion and an average positional deviation of 133 m across the seven experiments. Sensitivity analysis indicates that the HHO inversion model exhibits certain robustness in estimating release rates, while the two-dimensional location of the release point is more susceptible to interference from noise in off-site monitoring data.

Abstract The long‐standing hypothesis that an ice sheet covered the Tibetan Plateau during the Last Glacial Maximum (LGM) has been refuted. Further research has indicated that the Plateau experienced more extensive glacial expansion during Marine Isotope Stage (MIS) 6 than during the LGM. However, whether the Tibetan Plateau hosted large‐scale glacier systems during MIS 6 remains an open question. Here, we selected the arid Qilian Shan of the northeastern Tibetan Plateau to reconstruct 90 m resolution glacier extents for three major glaciations over the past 190 ka. Integration of geomorphological and chronological evidence with a glacial model shows that glacial extent was least extensive during MIS 2, was intermediate during MIS 3b, and was most extensive during MIS 6. The modelled ice extents are consistently larger than previously reported. Specifically, an ice field with an area of ∼5.61 × 104 km2 likely developed on the western Qilian Shan during MIS 6.

Glaciers are vital freshwater resources in some arid regions, where glacier meltwater sustains downstream ecosystems, land-use activities and regional water security. However, systematic investigations of long-term glacier change and its hydrological implications remain limited in the Shule River Basin (SRB) of the Qilian Mountains, China. Our study quantified spatiotemporal glacier changes and their meltwater in the SRB from 1980 to 2100 by integrating Landsat imagery, digital elevation models, meteorological and hydrological data, existing glacier datasets and modeling. In 2025, 627 glaciers covering ~437.75 ± 35.46 km2 were identified in the SRB, representing a shrinkage of 213.2 km2 (mean 4.74 km2 a−1) from 1980 to 2025. The glacier surface elevation change rate reached −0.60 m a−1 during 2000–2014, with accelerated thinning after 2010. The mean glacier mass balance was −343.63 mm w.e. a−1 during 1970–2023 and glacier meltwater contributed an average of 25.1% to basin runoff during 2008–2021, peaking at 36.71% in 2011. Future projections under the SSP-119, SSP-245 and SSP-585 scenarios indicate continued glacier retreat, with peak glacier meltwater expected between 2024 and 2045. These findings highlight the declining hydrological regulatory capacity of glaciers and provide scientific support for sustainable land–water management and ecosystem resilience in glacier-fed arid basins.

Glaciers in the Qinghai-Tibet Plateau and its surrounding regions have been continuously retreating under ongoing climate warming. However, the rates and patterns of glacier change vary markedly among regions and even within the same region across different glacier sizes. The prevailing view holds that smaller glaciers shrink faster in area, yet glacier response to climate forcing involves both areal reduction and surface thinning. The spatial variability of surface elevation change across glaciers of different sizes remains insufficiently understood. Here, we analyzed 5907 glaciers across the Qinghai-Tibet Plateau and its surrounding regions using glacier inventories from the 1970s and 2018, in combination with glacier velocity datasets and PyGEM simulations. Between the 1970s and 2018, the total glacierized area decreased by 21.79% (from 16,632.80 km(2) to 13,007.79 km(2)). Glaciers smaller than 1 km(2) exhibited the fastest areal shrinkage (5.74% per decade), followed by glaciers of 1-5 km(2) (5.13% per decade) and 5-10 km(2 )(4.79% per decade), whereas glaciers larger than 10 km(2) showed comparatively smaller reductions (3.85% per decade). In contrast, larger glaciers (>10 km(2)) experienced stronger surface thinning (-0.53 m/y) than smaller glaciers (<1 km(2,) -0.43 m/y). Faster-moving glaciers (>3 m/y) underwent greater thinning but maintained relatively stable areas, while slower glaciers exhibited more pronounced areal retreat. Further analysis showed that the mean ratio of ice loss due to area shrinkage to that due to surface thinning (R-T/S) is 0.30 (median = 0.23), with a critical velocity of approximately 3.0 m/y. Glaciers with velocities below this threshold primarily lose mass through areal retreat, whereas those exceeding it are dominated by surface thinning. These findings reveal the distinct yet coupled roles of glacier size and velocity in determining glacier change characteristics, providing new insights for glacier monitoring, water resource assessment, and climate adaptation strategies in the Qinghai-Tibet Plateau and its surrounding regions.

Traditional source term inversion models rely on accurate a priori information as well as atmospheric dispersion simulations, leading to time-consuming source term inversion procedures. Previous studies have used machine learning (ML) methods such as neural networks to construct source term inversion models, which exhibit excellent inversion performance but usually lack model interpretability and have complex model structure and parameter tuning. To address this problem, an interpretable nuclear accident source term inversion model using ensemble learning combined with the SHapely Additive exPlanation (SHAP) method was developed in this study to estimate the nuclide release rate and the 2D location of the release point. In the model construction, Gaussian plume model is utilized to obtain data samples. To evaluate the adaptability of the model to accident scenarios, the model was trained under two types of accidents, known and unknown at the release point. The validity and accuracy of the model were assessed using statistical metrics, including the coefficient of determination (R2), root mean square error (RMSE), mean absolute percentage error (MAPE), and mean distance error (MDE). The CatBoost model showed the best performance in both scenarios compared to the other three models. Model feature importance calculations and SHAP analyses revealed that the radioactivity concentration monitoring data had the greatest impact on the model inversion performance in both scenarios, and wind speed was an important parameter for this inversion model. Variations in meteorological parameters critically impair the reliability of source term inversion under unknown release scenarios.

Surge-type glaciers are widely developed in mountainous areas around the world. Understanding the trigger mechanism of glacier surge is a prerequisite for addressing their impacts on hydrological assessments, disentangling climate-glacier linkages, and mitigating downstream hazards. Most glacier surges occur in the compound glaciers; however, attention paid to the trigger mechanisms of such surges is minimal. This study confirmed two surges in the northern and southern branches of the Aru-4 glacier, respectively, in the Western Tibetan Plateau, using multisource remote sensing data. The northern branch of the Aru-4 glacier entered the active phase in 1999 and the active phase lasted for 6 years. The southern branch of the Aru-4 glacier entered the active phase in 2007 and the active phase lasted for 9 years. The southern branch of the Aru-4 glacier experienced a long period of retreat before the northern branch surged and their tongues were in a detached state. The northern branch surge carried a large amount of ice to the frontal area, blocking the downward transport of ice from the southern branch and initiated surge. Through the analysis of two surge processes of Aru-4 glacier, we found a new surge mechanism for compound glaciers. It was revealed that surges in such glaciers are not only triggered by the reduction in basal sliding resistance caused by the internal factors. These surges initiated in the upper part of the glacier then propagated to down glacier by intense compression force. Furthermore, surges can also be triggered by external intervention of blocking by other branches. This external trigger initiates the surge in the lower part of the glacier then propagated to the upper part by longitudinal traction force. In addition, comparing with the surge triggered by the internal factors, the surge triggered by the external intervention may have a more dramatic process.

Source term inversion plays a critical role in consequence assessment during nuclear accidents. This study investigates three Bayesian approaches for machine learning-based source term inversion, with datasets generated by the radioactive nuclide atmospheric dispersion program RADC. Firstly, a Bayesian neural network (BNN) was designed using Python to predict the release rate of I-131 during nuclear accidents. The BNN achieved a mean absolute percentage error (MAPE) of 7.92%, showing significantly higher accuracy and robustness compared to a backpropagation neural network (BPNN) with an identical structure and under the same accident scenarios. Secondly, both BNN and BPNN are further enhanced through Bayesian optimization, achieving a significantly improved MAPE values of 5.78% and 3.80% respectively. Furthermore, The Monte Carlo Dropout method was used to generate confidence intervals for the BPNN, offering an uncertainty analysis approach for conventional neural networks in source term inversion. However, its uncertainty analysis showed greater fluctuations compared to BNN.

The inverse problems of the convection-diffusion equation (ICDE) have received extensive attention in incomplete boundary conditions and uncertain source terms. They can be applied in thermally stratified pipe elbows and so on. Many algorithms need to combine with optimization algorithms to repeatedly calculate the direct problem in the solution process. To solve such problems, this paper employs a boundary-type algorithm named the half-boundary method (HBM). The HBM does not require additional repeated optimization of the direct problem. To test the performance of the method, the numerical simulations of some problems have been carried out, including the inverse problems of heat convection, river pollution and air pollution. The results show that the HBM has the desired accuracy by comparing with the exact solution. If there are errors in the measurement process, the solution doesn't generate a large deviation from the result. It is worth noting that the placement of internal measurement points minimally impacts the numerical results within the solution domain. And the method is also able to handle with discontinuous problems. Because the Gaussian plume model verifies the accuracy of HBM, the HBM can quickly calculate the atmospheric diffusion of the non-Gaussian plume model.

The deuterium-tritium fusion represents a promising avenue for achieving commercial controlled nuclear fusion, but the usage of radioactive tritium poses significant environmental and radiological risks to personnel. To evaluate the radiological impact on nearby staff and the public from atmospheric tritium released during the steady-state operation of fusion devices, this study developed a dose assessment code based on the steady-state Gaussian plume model. Using tritium water vapor release rates from the ITER safety report as the source term, we calculated the tritium doses absorbed via inhalation and skin contact for various age groups, utilizing 2023 ECMWF reanalysis meteorological data for an inland basin site in China, a hypothetical site in Daya Bay, and the ITER site. The results indicate that the maximum public dose near the inland basin site is 0.030 mSv/yr, near the Daya Bay site is 0.015 mSv/yr, and near the ITER site is 0.018 mSv/yr, all of which are well below the ITER project's limit of 0.1 mSv/yr. Thus, fusion devices are radiologically safe for the surrounding personnel under normal operation.

Glacier-fed lakes are expanding, but research on the driving factors of their expansion is insufficient. The primary reason for this gap lies in the current definition of glacial lakes, which includes all lakes within 10 km of glaciers, regardless of whether they are glacier-fed or non-glacier-fed lakes, and does not consider glacier-fed lakes located more than 10 km from glaciers. To address this, this study presents a novel classification and analysis framework for glacier-fed lakes in High Mountain Asia using Landsat images and the CBAM-UNet model. We identify 17,382 glacier-fed lakes covering an area of 2,309.5 ± 223.4 km² as of 2020s. A key finding is that 21% of these lakes, accounting for 42% of the total area, are located more than 10 km from glaciers. Additionally, we highlight that the expansion of glacier-fed lakes is influenced by both glacier meltwater and precipitation. When the glacial contribution exceeds 10%, lake dynamics are predominantly driven by glacier changes. However, for lakes with a glacial contribution below 10%, precipitation plays a more significant role in shaping lake expansion. These findings offer new insights into glacier-lake interactions and provide a more accurate basis for future GLOF risk assessments in the region.

High-Mountain Asia (HMA) hosts the largest concentration of modern glaciers in middle- and low-latitude regions. The widespread glacial landforms in HMA suggest that these glaciers have experienced significant changes in extent over time. The climate of northern HMA is influenced mainly by the Asian monsoons and the midlatitude westerlies. Changes in the climate system during glacial-interglacial cycles potentially resulted in a unique pattern of glacial evolution in northern HMA; however, elucidating this pattern requires a comprehensive understanding of the spatiotemporal evolution of the glaciers in this region. To achieve this, we compiled 450 10Be exposure ages from northern HMA, including 84 new and 366 previously reported ages, and we also conducted high-resolution simulations of paleoglacier extent that correspond with glacial geological records. Our findings emphasize the complexity of past glacier evolution throughout HMA. The two climatic domains share similarities in glacial sequences, climatic mechanisms, glacial style, and depression of the equilibrium line altitude (ELA). Landforms created by glacier advances corresponding to Marine Isotope Stages (MIS) 6, 3, and 2 have been identified in both domains, indicating a strong relationship between these advances and cooler climatic conditions. Since the penultimate glaciation, glaciers in northern HMA have gradually evolved from extensive ice fields to piedmont glaciers, to valley glaciers, and then to cirque glaciers. This transformation is reflected in the progressive ascent of the ELA. A key difference in glacial evolution between these two climatic regions is that the timing of the local last glacial maximum (lLGM) was asynchronous compared to the global last glacial maximum (gLGM). The timing of the lLGM varied between the monsoons-westerlies transitional domain and the westerlies domain, occurring during MIS 3 and MIS 4, respectively. This difference was caused by variations in the relative strength of atmospheric circulation systems and changes in moisture supply. In a similarly cool climate, variations in precipitation distribution explain the differing patterns of ELA change observed in these two climate domains during past glacial cycles.

Sediment entrainment marks the initiation of particle motion on the bed surface and plays a crucial role in quantifying sediment transport. While the entrainment behavior might vary among different geophysical flows, the underlying mechanisms are often similar. To explore the shared dynamics, we compiled a global database of diverse geophysical flows: stream flow (SF), hyperconcentrated flow (HF), and debris flow (DF), obtained from field observations and laboratory experiments. We first validate existing Shield's number based entrainment frameworks but find them inadequate to account for entrainment fluxes of all flow types, particularly for the HF and DF, across a wide range of excess Shields number (theta/theta c) where theta c represents the critical value. Utilizing the random forest regression algorithm, we then proposed a new stream power (omega) dependent bursting area formula (A P) for all types of mass flows considered, resulting in a unified omega-based entrainment model. The revised model achieves an R 2 of 0.924, which is more than twice that of the theta-based model (R 2 = 0.427) when applied to the same compiled database. This work provides valuable insights for improving sediment transport modeling, which are essential for developing effective river management strategies and optimizing the design of related infrastructure systems.

Svalbard hosts numerous tidewater glaciers whose dynamic responses are critical to polar environmental change and global sea-level rise. This study investigates Tunabreen glacier and Osbornebreen glacier in Svalbard using multi-source remote sensing data from Landsat, Sentinel, ArcticDEM, and other reanalysis data. We analyzed glacier surface velocity, terminus length change, and elevation evolution to assess their dynamic characteristics and driving mechanisms. Tunabreen glacier experienced a typical surge from 2016 to 2018, advancing 1.6 km, reaching a peak velocity of ~2073 m yr−1 , and thickening by up to 15.7 m in the receiving zone. The surge propagated from the terminus upstream and was likely driven by subglacial hydrodynamics. In contrast, Osbornebreen glacier showed terminus advance and acceleration between 2017 and 2021, advancing 1.6 km with a peak velocity of ~3143 m yr−1 in 2020. However, the sustained thinning of the glacier, combined with fjord water warming driven by the North Atlantic Current and a significant increase in air temperature and precipitation, suggests that the observed acceleration was more likely externally forced by oceanic and atmospheric forcing. These contrasting behaviors reveal distinct response patterns of tidewater glaciers to internal glacier dynamics and external forcing, offering new insights into glacier evolution under climate change.

An accurate understanding of the effects of temperature/precipitation variations on geochemical and magnetic indicators within soils is fundamental to reconstructing the evolution of the Asian monsoon. Here, we investigate correlations between temperature/precipitation and geochemical/magnetic parameters in a Qilian Shan elevation transect. Our results suggest that the geochemical indicators of chemical weathering intensity are dominantly controlled by precipitation at low altitudes, but, at higher altitudes, temperature replaces precipitation as the primary controlling factor. In contrast, magnetic indicators consistently reflect precipitation influences across elevations. We explore this framework to address contradictions between magnetic and geochemical records from the Chinese Loess Plateau, proposing that late Neogene geochemical variations may reflect temperature shifts during global cooling, while magnetic changes align with precipitation, modulated by CO2 and tectonic paleogeography. We advocate for an integrated approach to reconstructing terrestrial temperature and precipitation histories.

Land use and cover change (LUCC) is the most direct manifestation of the interaction between anthropological activities and the natural environment on Earth's surface, with significant impacts on the environment and social economy. Rapid economic development and climate change have resulted in significant changes in land use and cover. The Shiyang River Basin, located in the eastern part of the Hexi Corridor in China, has undergone significant climate change and LUCC over the past few decades. In this study, we used the random forest classification to obtain the land use and cover datasets of the Shiyang River Basin in 1991, 1995, 2000, 2005, 2010, 2015, and 2020 based on Landsat images. We validated the land use and cover data in 2015 from the random forest classification results (this study), the high-resolution dataset of annual global land cover from 2000 to 2015 (AGLC-2000-2015), the global 30 m land cover classification with a fine classification system (GLC_FCS30), and the first Landsat-derived annual China Land Cover Dataset (CLCD) against ground-truth classification results to evaluate the accuracy of the classification results in this study. Furthermore, we explored and compared the spatiotemporal patterns of LUCC in the upper, middle, and lower reaches of the Shiyang River Basin over the past 30 years, and employed the random forest importance ranking method to analyze the influencing factors of LUCC based on natural (evapotranspiration, precipitation, temperature, and surface soil moisture) and anthropogenic (nighttime light, gross domestic product (GDP), and population) factors. The results indicated that the random forest classification results for land use and cover in the Shiyang River Basin in 2015 outperformed the AGLC-2000-2015, GLC_FCS30, and CLCD datasets in both overall and partial validations. Moreover, the classification results in this study exhibited a high level of agreement with the ground truth features. From 1991 to 2020, the area of bare land exhibited a decreasing trend, with changes primarily occurring in the middle and lower reaches of the basin. The area of grassland initially decreased and then increased, with changes occurring mainly in the upper and middle reaches of the basin. In contrast, the area of cropland initially increased and then decreased, with changes occurring in the middle and lower reaches. The LUCC was influenced by both natural and anthropogenic factors. Climatic factors and population contributed significantly to LUCC, and the importance values of evapotranspiration, precipitation, temperature, and population were 22.12%, 32.41%, 21.89%, and 19.65%, respectively. Moreover, policy interventions also played an important role. Land use and cover in the Shiyang River Basin exhibited fluctuating changes over the past 30 years, with the ecological environment improving in the last 10 years. This suggests that governance efforts in the study area have had some effects, and the government can continue to move in this direction in the future. The findings can provide crucial insights for related research and regional sustainable development in the Shiyang River Basin and other similar arid and semi-arid areas.