Affected by the severely cold climate， cold region tunnels are often faced with a series of frost damage problems， such as lining ice hanging， freeze-thaw damage or cracking， road icing and drainage system freezing failure， which pose great challenges to the construction and operation of tunnels. The sub-zero temperature environment is the necessary condition for frost damage in tunnels， thus mastering the longitudinal distribution of air temperature in tunnels is the basis and premise of frost damage research and engineering control measures in cold regions. However， various tunnels differ in environmental meteorology， engineering structure and traffic conditions significantly， resulting in obvious discrepancies in the longitudinal distribution characteristics of air temperature， causing many difficulties in accurately predicting and obtaining the longitudinal distribution of air temperature inside cold region tunnels. Therefore， in order to obtain the distribution characteristics and influence factors of air temperature in cold region tunnels， literature research and statistics and analysis methods were used to collect and sort the field measuring data and related influence factors of air temperature inside 52 cold region tunnels in China. According to the distribution characteristics of air temperature inside tunnels in cold regions， they are divided into three types： symmetric， asymmetric and cut-through type， and then the length and burial depth of each type of cold region tunnels are statistically analyzed. On this basis， the influential mechanism of temperature in tunnel site area， geothermal heat of surrounding rock and ventilation in tunnel on air temperature in the tunnels is discussed. Subsequently， the natural wind， engineering construction， mechanical ventilation and traffic factors that cause air flow inside the tunnel were analyzed， and then the primary and secondary influence order of each factor was pointed out. The research results indicate that when the natural environmental factors at both ends of the tunnel portal are basically the same and the elevation difference between the two ends of the entrance is not significant， the air temperature inside the tunnel usually shows a symmetrical distribution. When there is a significant slope inside the tunnel or it is affected by the prevailing wind direction outside the tunnel， the air temperature inside the tunnel often shows an asymmetric distribution. When the tunnel length is short， the burial depth is small and unidirectional wind is prevalent， the air temperature distribution inside the tunnel often presents a cut-through type. In addition， in terms of the primary and secondary factors affecting the temperature distribution inside cold region tunnels， the environmental temperature in the tunnel site area（tunnel entrance and exit） plays a decisive role in the distribution of temperature inside the tunnel， the burial depth of the tunnel determining the geothermal effect has an important impact on the temperature inside the tunnel， and ventilation is the key factor affecting the temperature inside the tunnel. Among the various factors affecting ventilation， natural wind and longitudinal slope design are the main factors influencing the longitudinal distribution of air temperature inside tunnels with slight entrance and exit elevation differences and short lengths and tunnels with significant entrance and exit elevation differences and long lengths respectively， while engineering ancillary structures， mechanical ventilation and traffic factors are relatively secondary. This research will provide reference for the study of frost damage， insulation laying design and operation and maintenance of cold region tunnels.

Understanding the evolution history of rock glaciers is the key to reveal the response of periglacial environment to climate change. However， the chronology of rock glaciers is relatively inadequate due to the lack of suitable dating materials and other reasons. Quaternary glaciation and periglacial environment were widely developed in Zheduo Mountain， located on the southeastern margin of the Qinghai-Xizang Plateau， which provided conducive conditions for the formation of rock glaciers. Based on the geomorphological investigation and ¹⁰Be exposure dating， we accurately dated a moraine-type rock glacier in Zheduo Mountain and further discussed its formation age and development characteristics. The results show a formation time of rock glaciers in the Zheduo Mountain at about （11.8±0.3） ka， corresponding to the late period of the Last Deglaciation. Millennium-scale climatic cooling events during the Last Deglaciation led to multiple cold and warm climate changes， leading to repeated glacier advances and retreats. The transport and accumulation of a large number of moraines strongly shaped the geomorphic landscape， promoted the continuous expansion of permafrost scale， and also provided the material basis for rock glaciers. Based on the studies of stone glaciers in the Qinghai-Xizang Plateau and other parts of the Northern Hemisphere， it is found that the formation age and evolution characteristics of the rock glaciers during the Last Deglaciation are similar. During the large-scale retreats of the glacier， sudden cooling events lead to the glacier advance and the expansion of the periglacial area， and when the temperature rises， the deglaciated proglacial area becomes a new periglacial environment， and glacial meltwater and sediment are generated at the same time， which promotes the development of the rock glaciers. The glacial and periglacial geomorphological processes， such as the repeated advance and retreat of glaciers， the expansion of permafrost areas， and the development of rock glaciers， caused by climate change， to a certain extent， indicate that the interaction of glacial and periglacial geomorphological processes in the interglacial cycle is the key to understand the response of the cryosphere to climate change.

Climate change has the potential to alter the global water cycle， resulting in an uneven spatial and temporal distribution of water resources. Furthermore， climate change can result in an increased frequency of meteorological and hydrological disasters. The Qinghai-Xizang Plateau is the most unique geological-geographical-ecological-resource-climatic unit on the Earth， known as “Asia Water Tower” and “The Third Pole”. It is the source of more than ten major rivers in Asia and is known as the “Water Tower of Asia”. With the intensification of global warming， extreme hydrological and meteorological events become more frequent on the Qinghai-Xizang Plateau. However， the complex climate and underlying surface characteristics， including glaciers， snow cover， and permafrost， limit our understanding of the response of extreme runoff to extreme precipitation in this region. Against this background， this study taking the Yarlung Zangbo River basin in the southern part of the Qinghai-Xizang Plateau as an example， this study， based on daily precipitation and runoff data， improves the method of identifying extreme streamflow and utilizes correlation analysis to investigate the relationship between extreme streamflow and extreme precipitation at different probabilities. The results indicate that extreme streamflow typically last 1~2 days in the Yarlung Zangbo River basin， and as the probability decreases， the duration of events correspondingly shortens. Extreme streamflow events with probabilities of 10%， 5%， and 1% exhibit coefficients ranged from 0.07 to 0.28 （P<0.01） with extreme precipitation， and the relationship strengthens with increasing probability of extreme precipitation. At the Nuxia Hydrological Station， the response intensity of extreme streamflow to extreme precipitation in the upper stream of Nuxia Hydrological Station is greater than that at the Nuxia and Linzhi Meteorological Station， with a lag of one day. Additionally， the response of extreme streamflow to extreme precipitation is also influenced by moisture content， vegetation conditions， and soil moisture changes in the Yarlung Zangbo River basin. The improved methods of identifying extreme streamflow in this paper can effectively capture the extreme streamflow events influenced by extreme precipitation in the Yarlung Zangbo River basin， and improve the understanding of the relationship between extreme streamflow and extreme precipitation response in high-altitude basins with complex streamflow sources. At the same time， this paper offers significant theoretical support and scientific guidance for the management of water resources and the regional economic development of the Yarlung Zangbo River basin in the context of climate change.

Against the backdrop of global climate warming and the warming effects induced by human activities， the permafrost on the Qinghai-Xizang Plateau is experiencing accelerated degradation， leading to frequent occurrences of retrogressive thaw slumps. This phenomenon not only directly impacts the regional ecological environment， but also poses a potential threat to the stability of engineering structures. This paper aims to provide a scientific basis for the prevention and prediction of thawing hazards and the assessment of climate change by reviewing the research progress on retrogressive thaw slumps in the permafrost regions of the Qinghai-Xizang Plateau by summarizing the formation mechanism， extraction methods， distribution characteristics， influencing mechanism， as well as the subsequent impacts of retrogressive thaw slumps on environment and engineering. It is concluded that the formation of retrogressive thaw slumps is the result of the interaction of various complex factors and long-term accumulation， often resulting in headward erosion， with diverse development shapes mainly controlled by slope gradients. They present various developmental shapes， mainly including elongated， branched， and multi-headed tongue shapes. Spatially， retrogressive thaw slumps show inhomogeneous distribution， which cluster in engineering corridors on the Qinghai-Xizang Plateau. Their spatial distribution is mainly controlled by terrain and permafrost conditions. Furthermore， retrogressive thaw slumps can alter soil structure and physicochemical processes， affecting regional carbon cycling， thereby increasing global or regional greenhouse gas emissions. Based on high spatial resolution remote sensing images interpretation and unmanned aerial vehicle （UAV） field verification， retrogressive thaw slumps was widespreadly monitored. Currently， research often combines Planet Cube Sat imagery with deep learning algorithms to achieve automatic identification and mapping of retrogressive thaw slumps in large-scale permafrost regions. Deep learning models such as Deep Lab V3+ show a better performance in future extensive studies of retrogressive thaw slumps. With the increasing demand for engineering construction and ecological environment protection on the Qinghai-Xizang Plateau， it is necessary to investigate the formation mechanisms， developmental characteristics， influencing mechanisms， and impacts on the environment and engineering of retrogressive thaw slumps， which will provide a theoretical foundation and scientific basis for engineering planning， prevention of retrogressive thaw slumps， and environmental protection on the Qinghai-Xizang Plateau.

Based on a review of existing studies and recent work， this paper collects and organizes research findings on glacier mass balance from 2013 to 2023. It summarizes the current methods and models commonly used in glacier mass balance research and reviews the status of such research in China. The results indicate the following： The commonly used methods for glacier mass balance research can be classified into three categories： traditional glaciological methods， geodetic imaging methods， and satellite gravity monitoring methods. Additionally， glacier balance models， which focus on the relationship between glacier mass balance （the algebraic sum of accumulation and ablation） and meteorological variables， have been developed. These models are based on energy balance equations and can be further divided into three subcategories： semi-empirical glacier balance models—degree-day models， energy-balance models， and coupled glacier flow and mass balance models， such as the Open Global Glacier Model （OGGM）. Under the influence of global warming， glaciers in China exhibit an overall negative mass balance trend with significant regional variations. Specifically： The No.12 Glacier in Laohugou， Qilian Mountains， experienced a severe mass deficit， with a cumulative loss of -71 760 mm w.e. between 1991 and 2020. The Kunlun Mountains initially showed relatively stable mass balance， but a negative trend has emerged in recent years. For example， glaciers in the eastern Malan Mountains lost -300 to -180 mm w.e. from 2000 to 2020. Glaciers in the Tianshan Mountains are significantly affected by temperature changes， showing a pronounced negative mass balance trend. For instance， glaciers in the Manas River basin lost -9 811.19 mm w.e. from 2000 to 2016， while the Urumqi Glacier No.1 lost -17 351.5 mm w.e. from 1956 to 2016. Glaciers in the Tanggula Mountains exhibit changes in mass balance similar to those in the Kunlun Mountains but with a more pronounced negative trend. Glaciers in the Dongkemadi River basin lost -7 550 mm w.e. from 1966 to 2015. In the Himalayas， glaciers show a slight negative mass balance. For example， glaciers in the Yamdrok Lake basin experienced a cumulative loss of -930 mm w.e. from 1987 to 2021. Glaciers in the Altai Mountains also exhibit a negative mass balance， with greater losses observed during 2000—2010 compared to the post-2010 period.

Scientific data has become a strategic resource， and scientific data centers， as crucial entities for the long-term storage and open sharing of scientific data， support the development of science and technology innovation. With an accumulation of scientific data construction spanning over 30 years， the National Cryosphere Desert Data Center （NCDC，http：//www. ncdc.ac.cn） aggregates over 60% of scientific data in the domestic glacial， permafrost， and desert research field. In order to promote the open sharing of scientific data resources in these field， NCDC has researched the construction of the resource systems， mechanisms， and methods for scientific data in cold and arid regions. The overall development objective is to build a scientific data system encompassing ice， snow， permafrost， deserts， and related elements， advancing theoretical methods for scientific data and formulate key technologies for the collection， transmission， storage， management， analysis， sharing， application， and product development， creating an integrated scientific data application resource pool that combines data， models， and computation. Simultaneously， assembling a group of interdisciplinary data teams to drive innovative work and establishing a benign development and sharing service ecosystem for scientific data. Additionally， advancing the disciplines of data engineering， key technologies for data engineering applications， and an international alliance for scientific data， constructing research-oriented， production-oriented， and intelligent scientific data centers with international influence， and providing scientific data and decision support for scientific and technological activities， ecological civilization， engineering construction， disaster prevention and control， and diplomatic negotiations in cold and arid regions.This paper reviews the development process and construction strategies of NCDC， summarizing its research practices in the field of scientific data sharing in cold and arid zones， covering aspects such as the construction of standard specifications， data resource management， system platform construction， and the effectiveness and progress of shared services. Finally， future development prospects are provided for the sustainable development of the National Scientific Data Center.

In seasonally frozen ground region landslides are become a more common geological disaster， which has attracted widespread attention in the world. Whether the physical processes of snow ablation and soil freezing and thawing in seasonally frozen ground region， as opposed to the non-monsoon freezing zone， have an impact on landslides deserves further study. A giant loess landslide group （Galamput landslide group） occurred on May 9， 2002 in the Ili region of China provides an ideal case study. By using field survey， remote sensing image identification， meteorological data analysis and loess characteristic test， we attempt to explore the formation process and failure mode， and further reveal the instability mechanism of Galamput landslide group. The results show that the Garamput loess landslide group is composed of three landslides with a total volume of approximately 17.355 million m³. The formation and development of the landslide group was a multi-stage and multiple sliding failure process. The Galamput landslide was the coupling triggering result of early snowmelt and post rainstorm. The snowmelt infiltration and soil freeze-thaw cycle in spring played an important role in the evolution process of slope deformation. The rapid infiltration of extreme rainstorm was the ultimate triggering factor of the landslide. In addition， special slope structures and stratigraphic combinations provide a material structural basis for the occurrence of loess landslides. Combined with the slope deformation process we established a deformation failure model for loess slopes considering the effects of precipitation infiltration and freeze-thaw cycles， and proposed that the combination of static liquefaction on the slip surface and sliding liquefaction at the foot of the slope is an important mechanism inducing the occurrence of loess landslides. In the future， the risk of large-scale landslides in the Ten-zan’s seasonally frozen ground region is extremely high with climate warming. This study can provide a new perspective on the formation process and failure mechanism of loess landslides， and is of great significance for understanding the early warning and risk assessment of landslide disasters in seasonally frozen ground region.

The Svalbard Archipelago is one of the most climate-sensitive areas in the world. Affected by the Arctic amplification effect， most of glaciers there have been undergoing significant shrinkage. Previous studies indicated that mass loss of glaciers in Svalbard accelerated in the past decades， but the ice mass loss caused by surge-type glaciers in the region remains unclear. Based on the high-precision observations from ICESat-2 laser altimetry， this study utilizes an improved classification method to investigate the elevation changes of the Svalbard’s glacier during the period from March 2019 to June 2022. The results show that， from 2019 to 2022， the elevation change of glaciers in the Svalbard shows a declining trend， with an average elevation change rate of （-0.94±0.23） m·a^-1， corresponding to a volume change rate of （-31.62±7.73） km³·a^-1. Among them， the surging glaciers occupy about 22% of the area， with a volume change rate of -13.23 km³·a^-1， accounting for 42% of the total volume change in the glacial area. Therefore， changes in surging glaciers are one of the main factors leading to the whole ice volume reduction in the region. As the largest surging glacier in the Svalbard， Storisstraumen Glacier has expanded its surge area by about 284 km² in the past 20 years， with a volume change rate of -5.67 km³·a^-1， accounting for 18% of the total ice volume loss in the area. Further analysis suggests that the increase in temperature may play a dominant role in triggering the surging of Storisstraumen glacier. This study not only reveals the elevation change characteristics of glaciers in Svalbard during the span from 2019 to 2022， but also quantifies the contribution from surge-type glaciers to the total ice mass loss in Svalbard， which may provide a reference for deep understanding the varying elevation change of glaciers in the region.

The headwaters of the Yellow River Basin （HYRB）， located in the northeastern part of the Qinghai-Xizang （Tibet） Plateau， serves as the main source of water and is highly sensitive to climate change. However， under the background of climate warming， understanding of shifts in precipitation types and the underlying causes remains limited. This study， focusing on this important scientific issue， utilized a parameterized model of wet-bulb temperature （a precipitation pattern recognition model using surface air temperature， surface pressure， relative humidity and elevation） to differentiate precipitation types across the HYRB from 1980 to 2015. Additionally， we analyzed changes in the annual mean precipitation， rainfall， snowfall and snowfall fraction during this period using the Mann-Kendall test and the Sen’ slope estimator. We especially focused on changes in snowfall fraction （SF， defined as snowfall/total precipitation） and analyzed spatial distribution of changes of SF. In order to reduce uncertainty in the differentiation of precipitation patterns by a single product， the analysis is based on weather station observations and multi-source high-resolution meteorological datasets， including the China Meteorological Forcing Dataset （CMFD）， the ECWMF ERA5-Land （ERA5-Land）， and the Multi-Source Weather （MSWX）. Further， we used the partial correlation analysis to conduct mechanism analyses of the underlying causes of long-term changes in SF in terms of both meteorological and circulation factors， including surface air temperature， relative humidity， precipitation， surface pressure， westerly winds， Indian monsoon， East Asian summer monsoon and ENSO. The results indicate that： （1） Annual mean SF exhibited significant （P<0.05） decreasing trends （0.002 a^-1）， because there was a decrease in snowfall， but rainfall increased significantly. The precipitation pattern shifted to rainfall. （2） Regionally， the percentage of grids with significant （P<0.05） decreasing trends in SF to the total grids of the HYRB was 98.31% for the ERA5-Land， 96.87% for the MSWX， and 67.62% for the CMFD. In particular， the west-central part of the HYRB exhibited a more rapid decline in SF， with Maduo （Madoi County） and Dari （Darlag County） experiencing the largest decreasing trends. （3） The three gridded datasets differed in their ability to capture the spatial variability of snowfall. Specifically， CMFD showed substantial spatial variation and presented well to spatial differences in trend of SF. MSWX was the second best， while ERA5-Land displayed a notable spatial homogeneity. （4） The annual mean SF was highly correlated with annual mean air temperature and precipitation （P<0.05）， with correlation coefficients r of -0.42 and -0.48. Moreover， westerlies also had a notable influence （r=-0.31， P=0.07）. Other factors had less effect on SF. These findings enhance our knowledge of the effects of climate warming on hydrometeorology. The identified trends and correlations offer theoretical support for the management of regional water resources in the Yellow River Basin and have important implications for water resources management across the HYRB.

Global warming has triggered significant changes in the cryosphere， manifested in phenomena such as glacier retreat， snowmelt， and permafrost degradation. These transformations accelerate the conversion of solid water resources into liquid water， disrupting the long-term stability of water resource allocation in the Qinghai-Xizang Plateau’s cold regions. This paper takes the perspective of hydrological effects of cryosphere changes in cold regions， reviewing recent advancements in the understanding of hydrological processes under climate change in the Qinghai-Xizang Plateau. We analyze the current challenges and hotspots in hydrological research specific to the Qinghai-Xizang Plateau. Given that hydrological modeling is a crucial tool for studying the hydrological cycle， the structure and functionality of these models significantly influence the accuracy and direction of hydrological research. This paper summarizes the advantages and limitations of hydrological model algorithms for simulating glacier and snowmelt runoff in the plateau cold regions， and the characteristics of glacier， snow， and permafrost modules in 10 typical hydrological models. We also distill the main issues affecting the simulation accuracy of hydrological process models in this region. Our findings indicate that the limited availability of meteorological observation stations in the Qinghai-Xizang Plateau contributes to uncertainties in data input and parameter estimation. Moreover， a lack of comprehensive understanding of the intrinsic physical mechanisms governing hydrological processes in the cryosphere results in incomplete model structures， further impacting the simulation accuracy of these hydrological models. Finally， we discuss strategies for enhancing the accuracy of hydrological models through the integrated application of multivariate data and machine learning algorithms in cold regions.

The Arctic Monitoring and Assessment Programme （AMAP） Workgroup released a scientific assessment of “Mercury （Hg） in the Arctic” in 2021， which evaluated the contents and storages of Hg over Arctic environmental media and predicted the potential future changes of Hg contents in Arctic environment based on the observational and modelling studies of Hg in Arctic environmental media from recent years. The report stated that global atmospheric Hg emissions originated mainly outside the Arctic （>98%）， and some of which could be transported over long distances in the atmospheric circulation and deposited in Arctic ecosystems， where them participated in the biogeochemical cycling of Hg in the Arctic. The 2/3 of the Arctic atmospheric Hg was deposited to terrestrial environments and 1/3 to marine environments. Hg in terrestrial environments was mainly stored in soils， glaciers， and snow， and could be transported to the Arctic Ocean by riverine transport and coastal erosion. Hg could also be transported into the Arctic Ocean by ocean currents. Hg stored in the Arctic Ocean could be removed by escaping into the atmosphere， burying on the continental shelf and in deep-sea basins， and outflowing by ocean currents. The Arctic environmental Hg was predominantly in inorganic form and could be partially converted to methylmercury （MeHg） by microbial methylation processes in anaerobic environments. Reductions in primary emissions will have a greater impact on future changes of Hg contents in Arctic environment than climate change. And thus stringent and feasible global anthropogenic Hg reduction policies are needed to reduce the ambient Hg levels in the Arctic over the next 20 years. The report also noted that in the future， better quantification of local Hg releases from the Arctic， enhanced research and modelling of key environmental processes for Hg， studies of microbial communities carrying Hg methylation and demethylation genes， consistent projections of primary emissions and climate change， and global inventories of anthropogenic Hg emissions are needed to accurately assess Hg cycling in the Arctic environment and predict changes in Arctic environmental Hg levels.

Understanding the dynamics of glaciers is of paramount importance in the context of global climate change， as glaciers serve as sensitive indicators of environmental shifts. They are key components of the Earth’s freshwater reserves， influencing sea level， water resources， and the global climate system. The study of glaciers， therefore， extends beyond pure scientific inquiry； it has critical implications for predicting future water availability， assessing potential sea-level rise， and developing strategies for mitigating the impacts of climate change. Within this broader framework， glacier surface crevasses offer a window into the deeper processes at work within glaciers. These features are not merely physical anomalies but are indicative of the glacier’s internal and external dynamics. They provide insights into the stress and strain glaciers undergo as they move and deform， offering clues about the glacier’s velocity， the distribution of stresses within the ice， and the interaction between the glacier and its underlying bedrock. Moreover， crevasses influence the energy balance of a glacier’s surface by altering its albedo and affecting its meltwater dynamics， which in turn affects the glacier’s overall mass balance and movement. Thus， the detailed study of crevasses contributes significantly to our understanding of glacier mechanics and their response to climatic variations.This study focuses on the Yanong Glacier， located in the Kangri Karpo Mountain of the southeastern Qinghai-Xizang Plateau， an area noted for its complex terrain and significant glaciological interest. The deployment of Unmanned Aerial Vehicles （UAVs）， particularly the DJI M300 RTK UAV， has facilitated the acquisition of high-resolution orthophotos with a ground resolution of 0.03 m. This technological advancement enables the detailed mapping of glacier surface features， which is crucial for identifying and analyzing crevasse patterns and their distribution across the glacier’s surface.To address the challenges associated with crevasse detection， the study employs a sophisticated deep learning framework， marking a significant advancement over traditional analytical methods. The introduction of the Convolutional Block Attention Module （CBAM）-UNet model represents a novel approach in the field of remote sensing and glaciology. This model enhances the feature extraction capabilities of the conventional U-Net architecture by incorporating attention mechanisms that focus on relevant features for more accurate crevasse detection.The comparative analysis conducted in this study demonstrates the superiority of the CBAM-UNet model over established deep learning models such as U-Net， DeeplabV3+， PSPNet， and HRNet. Achieving a precision rate of 90.74% in identifying ice crevasses， the CBAM-UNet model proves to be exceptionally effective in the extraction of various crevasse types， including Transverse， Splaying， En Échelon， Marginal Crevasses， and Rifts. These crevasse types， indicative of the glacier’s dynamic behavior and influenced by the underlying topography， provide valuable information on the structural integrity and movement patterns of the glacier.The utilization of high-resolution UAV imagery in conjunction with advanced deep learning models offers a robust method for the detailed and accurate detection of glacier crevasses. This methodological approach not only enhances the precision of crevasse mapping but also facilitates the continuous monitoring of glacier transformations. Such advancements are instrumental in understanding the complex responses of glaciers to climate change， contributing essential knowledge to the fields of glaciology and climate science.Furthermore， this investigation sheds light on the potential applications of UAV technology and deep learning in glaciological research， highlighting the importance of interdisciplinary approaches in addressing environmental challenges. The findings of this study underscore the utility of combining high-resolution remote sensing data with machine learning techniques to improve the accuracy and efficiency of glacier monitoring efforts.In conclusion， the research presented herein exemplifies the significant contributions of technological innovations to the study of glacier dynamics. By enhancing our ability to accurately map and analyze glacier surface crevasses， this study offers valuable insights into glacier behavior， the impact of climate change on glacier systems， and the broader implications for water resources and sea-level rise. It is hoped that the methodologies and findings of this study will serve as a foundation for future research in glaciology， promoting further advancements in the understanding and preservation of these vital natural resources.

The young glacial erosional and depositional landforms beyond the modern glaciers on the Qinghai-Xizang Plateau and its surroundings are direct imprints of the glacier fluctuation during the past several decades or century， and contain important palaeoclimatic and palaeoenvironmental information. Studies of these landforms could provide insights into the spatiotemporal variations of these ancient glaciers. Improved logistics and the development and application of the terrestrial in situ cosmogenic nuclide （TCN） ¹⁰Be surface exposure dating techniques， are making it to be possible to investigate the young glacial landforms in the Mount Geladandong area， which locates in the west segment of the Tanggula Mountains， central Qinghai-Xizang Plateau. Here we report on ¹⁰Be surface exposure dating of six samples collected from modern glacial landforms associated with the Gangjiaquba Glacier on the eastern slope of Mount Geladandong. Dating results demonstrate that the boulders ［samples from G19-02 to G19-06 and their ages are （125±15） a、（65±13） a、（86±13） a、（96±15） a、（104±15） a］were almost not affected by nuclide inheritance， and the ¹⁰Be surface exposure ages could be used to constrain glacier fluctuations during the past century in this high-altitude area. However， a comparison of the ages of a glacially polished surface ［sample of G19-01 and its age is （11 076±688） a］ and boulders indicates that subglacial erosion by abrasion of the rock surface was insufficient （<2 m） to remove cosmogenic nuclide inventories. Thus， the apparent exposure age on the polished surface includes inheritance and overestimates the true age of this surface. Our study suggests that a cautious sampling and interpretation strategy should be adopted with bedrock samples， and the ¹⁰Be surface exposure dating techniques have potential applications for dating glacier fluctuations on a 10¹-year time scale in these high-altitude areas with high ¹⁰Be production rates. Meanwhile， this study also provides a new reference for the application of ¹⁰Be exposure dating technique in dating the young landforms in high altitude areas.

The active layer， which is the primary area for permafrost and atmosphere to exchange heat and water， and its morphological changes have a significant impact on ecosystems， topography and geomorphology， surface and subsurface hydrological processes， carbon cycle， and safety of engineering structures in cold climates. In this paper， the spatial distribution characteristics of active layer thickness in the Three River Source Region （TRSR） and its future changes are studied based on the modified Stefan model， and the factors that may affect the variation of active layer thickness are discussed. The results show that the thickness of active layer has been increasing at a rate of 7.2 cm∙（10a）^-1 from 1981 to 2018， and will continue to increase in the future， most significantly in the central part of the TRSR. From 2006 to 2049， the active layer thickness increases at a rate of 4.3 cm∙（10a）^-1 （RCP6.0） to 6.8 cm∙（10a）^-1 （RCP8.5）； from 2050 to 2099， it increases at a rate of 0.04 cm∙（10a）^-1 （RCP2.6） to 4.8 cm∙（10a）^-1 （RCP8.5）， which is significantly less than the rate of increase in the previous 50 years. The rate of increase in active layer thickness has slowed down. Analysis of the factors influencing the variation of active layer thickness shows that the greatest influence is on mean annual air temperature， while the influence of annual precipitation is not significant， and NDVI has some influence on active layer thickness. This result will have an important impact on the water conservation， climate regulation， and ecological security of the TRSR.

Surface energy balance is an important part of the land-atmosphere interaction， which is of great significance for the study of energy exchange， water cycle， climate change and biodiversity. The Qinghai Lake watershed is an important barrier for the ecological security of the Qinghai-Xizang Plateau， and the swamp wetland is the main ecosystem type in the basin. However， there has been little research on the energy balance of swamp wetlands in the Qinghai Lake watershed. Based on this， this study analyzed the variation characteristics of surface energy balance of alpine swamp wetland in Qinghai Lake watershed by using the observed data obtained from the Wayanshan site in 2019. The results show that the surface radiation flux of the swamp wetland at the Wayanshan site has seasonal variation characteristics， and the downward and upward shortwave radiation values are highest in spring and lowest in winter. The net radiation， upward and downward longwave radiation fluxes are the largest in summer， followed by spring and autumn， and the smallest in winter. The net radiation is mainly consumed by latent heat flux and sensible heat flux， accounting for 60.6% and 35.3% of the annual effective energy， respectively. The effective energy is mainly consumed by latent heat flux in summer and autumn， and by sensible heat flux in winter and spring. Compared with other underlying surfaces of the Qinghai-Xizang Plateau， the surface latent heat flux consumes a larger proportion （>75%） of the effective energy in the swamp wetland in summer and autumn. The annual surface energy closure rate of swamp wetlands was 0.69， and the closure rates during the freezing and non-freezing periods were 0.51 and 0.74， respectively.

Glacial lake outburst floods （GLOFs） are a unique type of natural hazard in the cryosphere that may result in catastrophic fatalities and damages. The Himalayas and Nyainqentanglha are second in the world regarding the frequency and severity of the GLOFs after the Peruvian Andes. Slope movements such as ice avalanches， landslides， and rockfalls have been blamed for most GLOF triggers. Despite the intuitive simplicity of the idea behind catastrophic erosional incision by a large wave， this explanation has many possible difficulties. The moraine-dammed lakes are shallow-long lakes， and the solitary displacement wave is the only type wave may be produced by massflows. The overtopping of the moraine dam may be only the first solitary displacement wave because the amplitude damping of the wave is more than 63% during two reflections if the losses of reflections were not considered. However， the erosion is usually a slow process on the hydrodynamic timescale that characterizes the passage of a displacement wave， it is not sufficient that erosion could occur during the relatively fast traversal of the dam. The erosion caused by a single wave was too small to breach the dam. However， it was far easier for a constant water flow to breach the dam by overflowing than for erosion to breach the dam by wave overtopping. A giant wave must be generated by a giant volume of ice avalanches or landslides entering a lake， and result a large water level increasing. The overflowing high water level was the reason for the breaching of moraine dams. Six moraine lakes， drained or undrained， were investigated in Xizang， China. The particle size of armor layers on the surface and the slope angle of these moraine dams and their spillway were measured. A series of laboratory experiments on armor initiated by overflow in the spillway of the moraine dam were conducted. The four statuses of moraine lake were simulated during the experiments. The moraine dam drained when the armor was removed by overflow. The initial model and the critical water depth of the moraine dam breaching were obtained. The initial armor models were the fluvial transport and bed failure models according to the bed slope range. The initial armor model was the fluvial transport model when the bed slope was less than 19.6°. The bed failure model occurred when the bed slope was over 25.2°. When the bed slope was in the range of 19.6°~25.2°， the initial model was a mixed regime. The depth of overflow on the armor and the gradient differences between the upstream and downstream of the knick point were the key parameters to initiate the armor in the fluvial transport model. The relative degree of exposure at the knick point was less and resulted in the first coarse particle initiated by the fluvial transport model occurring at the knick point. The larger the gradient difference between the upstream and downstream of the knick point， the smaller the relative degree of exposure value was. The ratio of the moraine dam’s critical water depth to the armor diameter was inversely proportional to the gradient differences between the upstream and downstream of the knick point in the fluvial transport model. For the bed failure model， the depth of overflow on the armor， the internal friction angle of saturated armor， the bed slope of armor， and the saturated density of the armor were the key parameters to initiate the armor. For the bed failure model of the armor initiation， the ratio of the moraine dam’s critical water depth to the armor’s diameter was a fixed value. The moraine dam breaching prediction model was validated reasonably well by the field data in Xizang， China， and British Columbia， Canada.

In the water cycle， water bodies show different characteristics of stable hydrogen and oxygen isotopes （δ¹⁸O and δ²H） in different processes of evaporation， transport， convection， and condensation due to the influences of isotope fractionation. Therefore， δ¹⁸O and δ²H is widely used in the study of paleoclimate and modern hydrological processes. Previous studies mainly focused on the variations of precipitation stable isotopes in the low-altitude regions in the Lhasa River basin， a critical area for the studies of the progressing and evolution of monsoon and westerly wind systems. In contrast， studies on δ¹⁸O and δ²H data obtained from the alpine regions are largely lacking. In this study， we analyzed 347 event scale precipitation samples collected at three sampling sites in the Kuoqionggangri Glacier region from July 2020 to July 2023. The spatial and temporal variations of precipitation δ¹⁸O， the local meteoric water lines， the relationship between precipitation δ¹⁸O and meteorological factors， and the relationship between precipitation δ¹⁸O and convective activity are investigated to understand the influences of the Indian monsoon and westerly wind on the precipitation δ¹⁸O and δ²H in the Kuoqionggangri Glacier region at the source of the Lhasa River of the southern Qinghai-Xizang Plateau. Besides， the backward trajectory of water vapor was further demonstrated through correlation analysis and cluster analysis， so as to reveal sources of water vapor. The results showed that there was little difference in temperature， relative humidity， and precipitation among the three fixed points located in different altitudes in this study area from July 2020 to July 2023 （the glacier terminus （5 544.5 m a.s.l.）， the basin source （5 374.0 m a.s.l.） and the basin export （4 941.3 m a.s.l.））. In addition， the precipitation δ¹⁸O and local meteoric water lines were similar among these three sampling sites from July 2020 to August 2020. This suggested that the climate conditions remained relatively identical within the Kuoqionggangri glacier region. According to the above results， we put emphasis on data of the precipitation δ¹⁸O collected at the basin export from July 2020 to July 2023. The results revealed a two-stage pattern of changes in daily precipitation δ¹⁸O： a higher value before mid-June followed by a lower value. The monthly precipitation δ¹⁸O shows the highest value in June and the lowest value in September. The slope and intercept of the local meteoric water line during the monsoon period （8.12， 11.78） were obviously smaller than those during the non-monsoon period （8.79， 23.18）， indicating that the water vapor source of the precipitation during the monsoon period possessed a higher relative humidity compared with that during the non-monsoon period. The slope and intercept of the local meteoric water line around the year （8.27， 15.10） were more similar to those in the monsoon period. This phenomenon might be due to the large contribution of precipitation in the monsoon period around the year in this region. Monthly precipitation δ¹⁸O exhibited a significant dependence on temperature in the monsoon period. Specifically， the monthly precipitation δ¹⁸O and temperature are positively correlated. Daily precipitation δ¹⁸O were significantly dependent on precipitation amount around the year. Specifically， the daily precipitation δ¹⁸O and precipitation amount are negatively correlated. The convection activity taking place 1~6 days before the precipitation event would deplete the precipitation δ¹⁸O. This influence on the precipitation δ¹⁸O was mainly concentrated in the monsoon period. Results of the backward trajectory tracking analysis indicated that water vapor transported by the Indian monsoon contributed the most to the precipitation of the region throughout the year， which depleted the precipitation δ¹⁸O. This study preliminarily reveals the spatial and temporal variations of precipitation δ¹⁸O and its main influencing factors in the alpine mountains of the southern Qinghai-Xizang Plateau. Results of the current work can provide basic data for the study of water cycle in the alpine regions.

The evolution of the East Antarctic Ice Sheet can significantly affect global climate change since deglaciation could contribute much to sea-level dynamics. Thus， studying ice sheet history in this region is the key to understanding the Quaternary global and regional climate change. In recent years， in-situ Terrestrial Cosmic Nuclides （TCN） exposure dating has been gradually applied to the dating of glacial dynamics in the Larsemann Hills area. The data of Quaternary glacial chronology in this area can be employed further to understand the changes in the Antarctic Ice Sheet， and the dating results from this method currently require further refinement. This study summarized and homogenized 196 ¹⁰Be exposure dating data from 1997 to 2022 of the Larsemann Hills and adjacent areas. These areas include the Larsemann Hills， Vestfold Hills and Rauer Group， Prince Charles Mountains， and Grove Mountains. The sampling information such as sample type， altitude， latitude， elevation， shielding factors， thickness， erosion rate， density， and AMS results were also counted. Based on distinguishing erratic and bedrock samples， we applied the newest model to all of the sample data and recalculated them. Meanwhile， the onset of biogenic sedimentation in the Larsemann Hills has been recorded with bedrock exposure data for further discussion. The results showed that the minimum exposure age of the Larsemann Hills was （4.05±0.81） ka， and the maximum exposure age was （147.01±11.80） ka. The minimum age of exposure in its adjacent areas ranged from （0.32±0.20） ka to （4 096±2 404） ka. The earliest date reaches the mid-Pleistocene， specifically MIS Gi24， spanning the entire Quaternary period. Among all the samples discussed， 17.86% have an exposure age of less than 11 ka， 83.16% have an exposure age of less than 600 ka， and 98.47% have an exposure age of less than 2.8 Ma. We divided the chronology into the Holocene （approximately 11 ka， MIS 1）， the mid to late Pleistocene （approximately 11 ka to 600 ka， MIS 1 to MIS 15）， and the early Pleistocene to late Pliocene （600 ka to 4 000 ka， MIS 15 to MIS Gi24）. We reviewed and summarized the glacial evolution history of the Larsemann Hills and adjacent areas over these periods. For the samples from the Larsemann Hills， we conducted a detailed discussion on the impact of erosion rates. We recalculated the exposure ages of the samples using erosion rates of 0 mm·a^-1， 0.0007 mm·a^-1， and 0.0015 mm·a^-1 （modern erosion rates）. The results showed that erosion rates lead to a significant overestimation of exposure ages， and using modern erosion rates for correction resulted in model saturation. Finally， we discussed the relationship between exposure age and altitude by region. There was a clear pattern in the Prince Charles Mountains and Grove Mountains （alpine setting） but in the Larsemann Hills， Vestfold Hills， and Rauer Group （coastal setting）， glaciers primarily retreated horizontally， showing no obvious pattern. We compared the multi-source homogenized data from different regions， and the following are reached： （1） 17.86% of the samples have exposure ages less than 11 ka， 83.16% of the samples have exposure ages less than 600 ka， and 98.47% of the samples have exposure ages less than 2.8 Ma. In some regions， there are still controversies regarding the ice retreat time determined by ¹⁰Be exposure dating. （2） The resultant increase amounts to 51.57% when the erosion rate reaches 0.007 mm·a^-1. When calculating exposure ages， it is important to consider reasonable erosion rates. （3） The exposure age tends to decrease with decreasing altitude. However， in low-altitude coastal areas， this trend may be offset by various factors， which require further research to explain. Future dating work in this area should be further promoted， and attention should be paid to erosion rate selection.

In the context of climate change and intensive human activities， the water environment problem of watershed is still one of the severe challenges faced by watersheds management in China and abroad. The Taohe River basin is a typical alpine ecologically fragile basin in China， and the study of changes in water environment elements and their driving mechanisms in Taohe River basin is of great significance to the management of water environment in Taohe River basin. In this paper， the characteristics of precipitation and temperature change trends and mutation features in the Taohe River basin during the period 1990—2018 were analyzed by statistics， linear fitting and M-K trend and mutation tests. The land use transfer matrix was used to analyze the land use change characteristics of the Taohe River basin in 1990 and 2018， and the hydrological and meteorological dynamic characteristics of the Taohe River basin were revealed. On this basis， the SWAT hydrological model database was constructed by DEM， land use， soil and meteorological data， and the monthly scale simulation of runoff and water quality in the Taohe River basin was carried out. The model divided the Taohe River basin into 24 sub-basins. The meteorological data of temperature， precipitation， relative humidity， wind speed and solar radiation were input， and the weather generator was constructed by SWAT-Weather. The runoff simulation period was 1988—2018， and 1988—1989 was set as the preheating period. Due to the limited measured data， the DO simulation period is 2015—2018， and the NH₃-N simulation period is 2018. The simulation calibration and verification were carried out by using the SUFI-2 calibration method in SWAT-CUP， and the R²， NSE and PBIAS indexes were selected to evaluate the simulation accuracy of the model. In order to improve the accuracy of the simulation， 22 runoff-related parameters and 37 water quality-related parameters were selected. Based on the simulation results of SWAT model， the temporal and spatial variation trend of runoff in Taohe River basin is analyzed by using water yield index and Slope trend. Finally， by setting different input conditions of precipitation， temperature and land use type， the scenario simulation was carried out for climate change， land use change and comprehensive change of climate and land use， and the driving mechanism of runoff and water quality in Taohe River basin is analyzed based on the scenario simulation results. The results show that： （1） During the past 30 years in the Taohe River basin， The interannual variation of precipitation fluctuates greatly， and the overall trend is upward， with a linear trend of 20.16 mm⋅（10a）^-1； There is no significant mutation in temperature， and the overall trend is upward， with a linear trend of 0.55 °C⋅（10a）^-1. The land use types in the Taohe River basin are mainly agricultural land， grassland and forest land. All land use types have changed， but the area changed slightly. （2） The SWAT model has good applicability to simulate long-term monthly runoff in the Taohe River basin； the R² and NSE are generally above 0.6， and the absolute value of PBIAS remains within 15% The accuracy of monthly scale short-term simulation of DO is relatively high， and the accuracy of monthly scale short-term simulation of NH₃-N is slightly lower， but the results in the middle and lower reaches meet the simulation accuracy requirements. The water yield in the upper reaches of the Taohe River basin is the highest， and the trend of increase is larger. The water yield in the middle and lower reaches is negatively correlated with the trend of increase in water yield. （3） Runoff is positively correlated with and controlled by precipitation， but negatively correlated with temperature. The runoff yield of each land use is agricultural land > grassland > forest land. The concentration of NH₃-N is negatively correlated and controlled by temperature， but positively correlated with precipitation. The conversion of agricultural land to forest land and agricultural land to grassland significantly reduce the concentration of NH₃-N. The concentration of DO is negatively related to temperature. and is almost unaffected by the other factors.

In order to explore the characteristics of winter temperature changes and extreme cold weather in Heilongjiang Province， cope with extreme climate events， daily temperature data and ERA5 reanalysis data from 84 national meteorological stations in Heilongjiang Province during 1961/1962—2022/2023 were selected， and the spatio-temporal distribution characteristics of multi-year temperature in Heilongjiang Province were analyzed by means of synoptic and statistical methods. the atmospheric circulation during extreme temperature was discussed， and the atmospheric circulation index which had better indication of winter temperature was selected. The results show that the winter mean temperature and extreme minimum temperature in Heilongjiang Province are higher in the south and lower in the north， and both show a warming trend， and the extreme cold weather shows a decreasing trend， but the longest duration and maximum influence range of the extreme cold weather have no obvious decreasing trend. The extreme cold weather mainly occurred in the north of Heilongjiang Province， where the Greater Khingan Mountains accounted for 68.7% of the total extreme cold days， followed by the Lesser Khingan Mountains 27.6%， and other regions only 3.7%. Most of the extreme cold weather occurred in January， accounting for 56.9% of the total number of extreme cold weather in winter， followed by December （24.5%）， and February （18.6%）. The abrupt changes in winter mean temperature and extreme cold station occurred in the 1980s， while the abrupt changes in extreme minimum temperature occurred in the 1990s， which are more consistent with the abrupt changes in the Arctic Oscillation， the East Asian trough intensity， and the Arctic sea ice area in autumn. The winter temperature in Heilongjiang Province is mainly affected by the polar circulation， and the southerly polar vortex may lead to a long duration of extremely cold weather in Heilongjiang Province. The Kuroshio Current SST Index， East Asian Trough Intensity Index and Scandinavian Pattern have good correlation with winter temperature in Heilongjiang Province， which have good indication for predicting future extreme temperature changes.

The warming of the Qinghai-Xizang Plateau has led to glacier retreat， permafrost melting， and an increase of meteorological and derived disasters， capturing widespread attention across society. In recent years， the warming of the Qinghai-Xizang Plateau has intensified further. In 2022， the summer was the warmest since 1961， with the average summer maximum temperature （T_max） and minimum temperature （T_min） in the central and eastern Qinghai-Xizang Plateau being 2.37 ℃ and 2.51 ℃ higher than that during the period of 1961—1990. The attribution technique， combining model evaluation and reconstruction， was employed to detect and analyze the anthropogenic influence on the extreme temperature events of the summer 2022 over the Qinghai-Xizang Plateau， utilizing CMIP6 model simulation data. The results indicated that greenhouse gases emissions from human activities significantly heightened the probability of maximum temperature extreme events during the summer 2022 over the Qinghai-Xizang Plateau. The probability of extreme T_max events occurring with and without the influence of human activities was 3.67% and 0.012%， respectively. The contribution of human activities to extreme T_max events in summer 2022 was estimated to be 1.26 ℃ （90% CI： 0.86~1.68 ℃）. The probability of extreme T_min events occurring with and without human activities is 23.5% and 0， respectively. The contribution of human activities to extreme T_min events in summer 2022 was estimated to be 2.35 ℃ （90% CI： 1.89~2.81 ℃）. The CMIP6 model underestimated the observed temperature changes over the Qinghai-Xizang Plateau. The simulation deviation of the model is calibrated based on the attribution constraint method， and the projected summer maximum and minimum temperature over the Qinghai-Xizang Plateau are expected to continue to increase in the future under the medium emission SSP2-4.5 scenario of shared socioeconomic path. The risk of extreme T_max and T_min events similar to those in 2022 occurring on the Qinghai-Xizang Plateau in the future is increasing.

As a sensitive region to global climate change， the climate of the Tibetan Plateau has gradually changed to a warm and humid state in recent decades， and this change and the resulting permafrost changes have directly affected the hydrothermal conditions for vegetation growth. The normalized difference vegetation index （NDVI） is an important variable to characterize the growth of vegetation， and its time-series variation can well reflect the response of vegetation to natural factors. In order to more accurately characterize the impact of climate change on vegetation growth in the permafrost areas of the Qinghai-Tibet Plateau， this study selected the source of the Tuotuo River， where permafrost is relatively continuous， as the study area. Based on Landsat NDVI data， trend analysis， partial correlation analysis， and geographic detectors were used to analyze the response characteristics of NDVI in the vegetation growing season （May-September） in the study area from 2000—2021 to climate change and different terrains and vegetation types. The results show that： （1） NDVI in the growing season in the source area of the Tuotuo River showed an overall fluctuating growth trend during 2000—2021， with a growth rate of 0.013 （10a）^-1， of which the NDVI growth rate was faster during the period of 2000—2010 ［0.019 （10a）^-1］， and then slowed down to half of the growth rate of NDVI during the period of 2011—2021 ［0.011 （10a）^-1］. The interannual changes of NDVI in different vegetation types were analyzed， and it was found that the better the vegetation development， the faster the growth rate of NDVI， among which the fastest growth rate was found in alpine swamp meadow ［0.018 （10a）^-1］ and alpine meadow ［0.017 （10a）^-1］， followed by alpine grassland ［0.015 （10a）^-1］， and the slowest growth rate of NDVI was found in alpine desert grassland ［0.009 （10a）^-1］. The analysis of the inter-annual changes of NDVI in different terrain areas showed that the growth rate of NDVI increased and then decreased with the elevation， and the growth rate was faster between 4 700~5 300 m ［>0.02 （10a）^-1］；The growth rate of NDVI showed a trend of increasing first and then decreasing with the change of slope， and the fastest growth rate ［0.023 （10a）^-1］ was found in the gentle slope area with a slope of 2°~6°； the growth rate of NDVI decreased with the increase of the thickness of the active layer of permafrost， and the fastest growth rate ［0.029 （10a）^-1］ was found when the thickness of the active layer was less than 150 centimeters. （2） On the spatial scale， the areas with an increasing trend of NDVI in the growing season in the source area of the Tuotuo River during the period of 2000—2021 accounted for a total of 82.1% of the total area of the study area， and the area with a highly significant increasing trend accounted for 27.0%， which was mainly distributed on both sides of the river at lower elevations and in the areas near the glacier； and the areas with decreasing trend of NDVI accounted for 16.0%， which was mainly distributed in the northern part of the study area in the alpine desert grassland areas in the northern part of the study area. （3） From 2000—2021， the source area of the Tuotuo River showed an overall trend of warming and humidification， and precipitation， air temperature， solar radiation and soil water content all showed fluctuating upward trends in the growing season. 2000—2021 NDVI changes in the source area of the Tuotuo River in the growing season were most significantly affected by soil water content， and warmer temperatures would prompt the melting of the cryosphere and increase soil water content， which would improve the growth status of the vegetation； NDVI changes in the growing season showed the same trend as precipitation； NDVI changes in the growing season showed the same trend as precipitation. NDVI changes in the growing season were positively correlated with precipitation； the correlation between NDVI changes in the growing season and air temperature varied widely in space， with 55.4% of the regions showing a positive correlation and 43.8% showing a negative correlation； and NDVI changes in the growing season were negatively correlated with the solar radiation as a whole. Increased precipitation in the growing season in the source area of the Tuotuo River promoted vegetation growth in the meadow area more than in the grassland area； changes in air temperature promoted vegetation growth in the meadow area more than in the grassland area， and inhibited vegetation growth in the grassland area more than in the grassland area. （4） The spatial differentiation of NDVI in the study area is mainly influenced by soil water content， climate factors and slope， and the influence is greatest when soil water content and slope interact. This paper quantifies the effects of climatic factors and their interactions on vegetation changes using various methods through the study of NDVI changes in the source area of the Yangtze River， which further reveals the influence mechanism of regional climatic and environmental factors on NDVI changes， and provides a new basis for the study of vegetation changes and ecological environmental protection in the Tibetan Plateau region.

Ice thickness and glacier topography data are the basis of glacial dynamics simulation. The analysis of ice thickness distribution and terrain features is important to understand the characteristics of glacier flow velocity， stress and its change. In September 2021， a ground penetrating radar system was used to probe the ice thickness of the Ningchan River Glacier No.3 in the Qilian Mountains， then using Ordinary Kriging interpolation method to process ice thickness and glacier surface elevation data， and based on the processed data we analyzed the glacier’s profile and whole characteristics of ice thickness distribution and glacier topography， and also studied glacier’s changes in recent years. The results showed that： the thicker parts of the Ningchan River Glacier No.3 were mainly located in the areas where the bed terrain was relatively flat， and broadly the ice thickness value decreased from the center to the edge， and increased first and then decreased with the rise in altitude； compared with glacier surface， the bed terrain of the Ningchan River Glacier No.3 fluctuated more strongly， and the shape of glacier bed in each transverse section appeared in various forms， such as slope， double trough valley and V-shaped valley； the glacier area， maximum ice thickness， mean ice thickness and ice volume of the Ningchan River Glacier No.3 in 2021 were respectively about 1.08 km²， 60 m， 24.1 m and 0.026 km³， and the glacier area， mean ice thickness and ice volume had respectively reduced about 0.123 km²， 3.3 m and 0.007 km³ from 2009 to 2021， and their mean annual change rate were respectively about -1.03×10^-2 km²·a^-1 （-0.86%）， -0.28 m·a^-1 （-1.00%） and -5.83×10^-4 km³·a^-1 （-1.77%）， compared with the period 1972—2009， the glacier shrunk faster， and its primary cause was the rise of mean summer （June-August） air temperature； the change of ice volume was in the form of thickness thinning all the time from 1972 to 2021.

Ice and snow tourism， as a unique form of tourism， not only has significant economic value but also exerts profound impacts on regional ecological environments and social cultures. Due to its unique geographical and climatic conditions， Northeast China has become an important destination for ice and snow tourism in China. On the one hand， ice and snow tourism has brought significant economic benefits to Northeast China. It has promoted local economic development， increased employment opportunities， and improved infrastructure. This influx of tourism-related activities has stimulated various sectors， creating a ripple effect that boosts the overall economy of the region. The economic gains have allowed for better services， enhanced public facilities， and an improved quality of life for the residents. On the other hand， ice and snow tourism faces several challenges， including ecological and environmental pressures as well as social and cultural impacts. Without effective management， these challenges can lead to environmental degradation， resource depletion， and cultural homogenization. The natural landscapes that attract tourists are at risk of being damaged by overuse， and the local culture may become diluted as it adapts to cater to tourists. Currently， the development of ice and snow tourism in Northeast China is also encountering several difficulties and bottlenecks. One major issue is the regional imbalance in tourism development. While some areas， particularly those along the coast， have advanced rapidly， inland areas lag behind， resulting in uneven distribution of tourism benefits. Another significant challenge is the lack of scale efficiency， which hampers the overall effectiveness of the tourism industry in the region. Many areas struggle with unstable resource development， where inconsistent investment and planning lead to sporadic growth and underutilization of tourism potential. These problems not only constrain the sustainable development of ice and snow tourism but also affect the overall competitiveness of the region’s tourism industry. To address these issues， this paper uses the three-stage super-efficiency SBM model to measure the efficiency of ice and snow tourism development in Northeast China. Combined with the pressure-state-response theory， the panel Tobit model is used to analyze the influencing factors of efficiency. The main conclusions are as follows： （1） Analyzing the temporal characteristics， the development efficiency of ice and snow tourism in Northeast China from 2015 to 2023 shows an overall fluctuating upward trend. However， the main cause of inefficiency is identified as scale efficiency. Despite the upward trend， the fluctuations indicate periods of both progress and setbacks， reflecting the complex dynamics of tourism development in the region. （2） From the perspective of spatial characteristics， the efficiency of ice and snow tourism development in Northeast China from 2015 to 2023 exhibits a spatial distribution pattern of “decreasing from coastal to inland areas”. This spatial gradient highlights significant regional disparities in tourism development. The formation of two major ice and snow tourism circles centered around Shenyang and Harbin underscores the concentration of tourism activities and resources in these areas. The average efficiency values follow a descending order from Liaoning Province， Jilin Province， Heilongjiang Province， to the eastern part of Inner Mongolia Autonomous Region. Most cities have not reached the ideal state， revealing prominent issues of regional imbalance and instability in the development of ice and snow tourism resources. （3） The results of the influencing factors calculation show that the development efficiency of ice and snow tourism in Northeast China is affected by various factors related to urban pressure， state， and response. Specifically， the pressure exerted by increasing tourist numbers and activities， the current state of tourism infrastructure and environmental conditions， and the response measures implemented by local governments and communities all play crucial roles. When developing the ice and snow tourism industry， it is essential to balance social pressures with ecological pressures. Strengthening social response mechanisms is vital to mitigate these pressures. This highlights the necessity of addressing regional disparities and promoting balanced development across different areas.

In the boreal permafrost region， forest fires not only affect the soil properties of forest ecosystems， but also the permafrost environment， soil organic carbon （SOC） and total nitrogen （TN） contents and storages. The permafrost region in Northeast China is located at the southern margin of Eurasia permafrost zone， which is undergoing extensive degradation under the influence of climate change and human activities. This affects carbon emissions， carbon and nitrogen storages in permafrost region， creating a positive feedback effect on climate warming. In this study， we selected the burned area in 2009 in Alongshan， northern Da Xing’anling Mountains， and the unburned site as a control to study the effects of different fire severity （light and severe burn） on the SOC and TN contents and storages in the permafrost region. The results showed that： （1） SOC and TN contents and storages varied significantly at different fire severities sites， and gradually decreased with increasing fire severity. Compared with the unburned site， SOC contents decreased by 9.78% and 65.11%， respectively； SOC stocks decreased by 9.29% and 68.48%， respectively； TN contents decreased by 1.99% and 52.49%， respectively； and TN stocks decreased by 3.23% and 51.61% at the light and severe burned sites， respectively. With the increase of soil depth， SOC and TN content gradually decreased； SOC storages showed a fluctuation pattern at burned site， increased first and then decreased at light burned site， and gradually decreased at severe burned site； TN storages showed a fluctuation pattern at unburned and light burned sites， and gradually decreased at severe burned site. （2） Soil temperature increased gradually with increasing fire severity. At depths of 0~100 cm， compared with the unburned sites， soil temperatures increased by （0.87±0.18） ℃ and （9.09±0.37） ℃ at light and severe burned sites， respectively； soil moisture contents （SMC） increased by （17.79±3.36）% at light burned site and decreased by （16.71±2.92）% at the severe burned site. Soil temperature and SMC decreased with increasing soil depth. （3） Redundancy analysis （RDA） showed that soil microbial biomass nitrogen （MBN） was the key factor affecting post-fire soil carbon and nitrogen contents and storages， with an explanatory rate of 65.6% （P=0.002）. SOC and TN contents were significantly positively correlated with soil temperature， SMC， NH₄⁺-N， microbial biomass carbon （MBC） and MBN， and significantly negatively correlated with fire severity， soil depth， and pH. SOC and TN storages were significantly negatively correlated with fire severity， soil depth， soil temperature， soil bulk density， and NO₃^--N. In summary， forest fire has a significant effect on SOC and TN storages， resulting in the loss of SOC and TN storages， altering the distribution pattern of carbon and nitrogen pools， and reducing the stability of soil carbon and nitrogen pools. The study of forest fires on SOC and TN contents and storages is of great significance to the ecosystem carbon balance and the management of ecological environment in the cold region， and it can provide important data references for the SOC and TN storages in the boreal permafrost region.

Snow cover is the main controlling factor of the freeze-thaw cycle of seasonally frozen ground in northern Xinjiang， and seasonally frozen ground affects the infiltration of snowmelt water by changing the freeze-thaw phase of shallow soil. However， the freeze-thaw state of shallow soil in the thawing snowmelt season in this region is not clear， making it difficult to accurately assess the synergistic regulation of snow cover and frozen soil on soil moisture from the mechanism level. Therefore， based on the monitoring data of snow cover and frozen ground at six meteorological stations in the Altai Mountains from 1961 to 2011， this study applied the Gaussian model and the Boltzmann model to analyze the basic characteristics of snow cover and seasonally frozen ground in northern Xinjiang on the basis of dividing high snowfall years， low snowfall years and normal years， and discussed the freeze-thaw state of shallow soil during the melting period in detail. The results show that the average duration of snow cover is 123.2 days and the average maximum depth of snow is 29.7 cm. The average annual freezing period of seasonal frozen soil is 150.9 days， and the average maximum freezing depth is 120.3 cm. In general， the snow cover showed an increasing trend， mainly manifested as an increase in snow depth. On the other hand， the frozen soil shows a degradation trend， which is mainly reflected in the shortening of the freezing period and the decrease in maximum freezing depth. The comparative analysis of the end time of the frozen soil melting and the end time of snow melt in different types of snow cover years shows that the end time of frozen soil melting in 70% snowy years and 60.5% normal years is 8.2 and 5.5 days earlier than the end time of snow melt， respectively. The end time of frozen soil melting in the year with low snowfall is 13.2 days later than the end time of snow melting. Overall， the results indicate that as snowfall increases， the probability of seasonally frozen ground being in a thawed state during the snowmelt period also increases significantly. The snowmelt water can recharge soil moisture， which leads to a longer retention time of snowmelt water in the soil. This snow-frozen soil synergistic mechanism significantly affects the infiltration of snowmelt water and alters the snow hydrology runoff process. This process can facilitate more snowmelt water replenishing the soil and strengthens the exchange between meltwater and groundwater， contributing to the effective utilization of snowmelt water in arid regions.

MODIS V6 no longer provides binary and fractional snow cover products， but only gives the Normalized Difference Snow Index （NDSI） of pixels. Therefore， when snow mapping based on MODIS V6， the selection of NDSI threshold and the corresponding snow mapping accuracy need to be studied. In this paper， based on the daily measured snow depth data of 250 ground meteorological stations from 2013 to 2021， we evaluate the NDSI_Snow_Cover band in 1927 scenes of MOD10A1 and 1936 scenes of MYD10A1 images in three typical snow cover areas in China. The optimal accuracy and corresponding optimal NDSI threshold of station-by- station pixels in snow cover product mapping were calculated respectively， and the factors affecting the accuracy were analyzed. The evaluation results based on station snow depth show that： （1） The mean ± standard deviation of the optimal NDSI thresholds of MOD10A1 and MYD10A1 were 0.16±0.09 and 0.17±0.10， respectively. The mean±standard deviation of the corresponding OA， FS and CK were 0.96±0.05 and 0.94±0.05， 0.84±0.19 and 0.75±0.24， 0.81±0.20 and 0.71±0.24， respectively. The accuracy of MOD10A1 was better than that of MYD10A1. （2） The optimal accuracy of MODIS snow cover products has obvious spatial heterogeneity. The accuracy in the Qinghai-Xizang Plateau is much smaller than that in the Northeast-Inner Mongolia region and the northern Xinjiang region. （3） In the station-based snow mapping accuracy evaluation， the threshold of snow depth will affect the evaluation results. The accuracy of snow mapping was the highest when the snow depth threshold of 2 cm and 4 cm was used to evaluate MOD10A1 and MYD10A1. （4） SCO and SDI have a significant positive correlation with CK of MOD10A1 and MYD10A1， and the correlation coefficients were 0.72 and 0.77， 0.67 and 0.71， respectively. （5） The terrain of the Qinghai-Xizang Plateau is complex， and the snow is mainly light snow. The snow information of the stations cannot represent the pixels well. Therefore， it is better to use synchronous UAV observation or higher resolution remote sensing images to evaluate the accuracy of snow cover products on the Qinghai-Xizang Plateau.

Lake ice phenology is the seasonal change of lake freezing， melting and other phenomena， which plays an important role in indicating regional climate change. This article is based on Google Earth Engine （GEE）， using 250 m spatial resolution MODIS surface reflectance products MOD09GQ and MYD09GQ， and Landsat series remote sensing data to analyze the lake conditions of 10 large lakes （area greater than 200 km²） on the Mongolian Plateau from 2000 to 2021. Ice phenology changes were studied， and their correlation with changes in temperature， precipitation， and Mongolian high pressure was analyzed using meteorological data. The results show that： （1） All lakes begin to freeze from November to mid-to-early December， and are completely frozen from mid-November to late December， entering the lake freeze period； they begin to melt in late April of the following year， and all lakes are completely frozen by the end of May. After thawing， the average lake freeze period for the 10 lakes is 151 days. （2） The overall phenology of lake ice in the study area shows the trend of delayed freezing date， advanced melting date， complete lake freezing period， and shortened lake ice existence period. The specific performance is that the average change rate between the day of initial freezing and the day of complete freezing is 1.88 d·（10a）^-1 and 1.91 d·（10a）^-1， respectively； the average rate of change on the day of initial thawing and complete thawing is -3.45 d·（10a）^-1 and -3.41 d·（10a）^-1； the average change rates of the lake ice existence period and complete freezing period are -5.32 d·（10a）^-1 and -5.37 d·（10a）^-1，respectively. （3） Climate change will affect the phenological changes of lake ice in the study area， and temperature is the key factor affecting the phenological changes of lake ice. Rising temperatures lead to a delay in the freeze-up start date （FUS） of lake ice， an advance in the break-up start date （BUS）， and a shortening of the complete freezing duration （CFD）. At the same time， the phenological properties of lake ice in some lakes will be affected by the Mongolian high pressure in winter， which will bring the cold air will cause the lake ice break-up duration （BUD）， lake ice cover duration （ICD）， and complete freezing duration （CFD） to increase.

Global climate warming causes permafrost to rapidly heat up and gradually degrade into seasonally frozen soil. There are significant differences between permafrost and seasonally frozen soil in terms of soil stability， water transport， and land-atmosphere interaction. Distinguishing between permafrost and seasonally frozen soil is essential due to their distinct hydrothermal dynamics， freeze-thaw behaviors， and sensitivities to climatic variables. This study is based on the coupled hydrothermal model （Simultaneous Heat and Water model， SHAW）， taking Dashalong Station （permafrost） and Arou Station （seasonally frozen soil） in the Qilian Mountains in the upper reaches of the Heihe River as the research objects， to simulate soil temperature， moisture and soil freezing and thawing processes. The results show that the SHAW model exhibits great accuracy in simulating hydrothermal processes at both types of frozen soil sites， but its overall performance is better at permafrost sites. Specifically， the average Nash efficiency coefficients of soil temperature and soil moisture at permafrost/seasonally frozen soil sites are 0.95/0.91 and 0.74/0.37. These findings underscore the SHAW model’s heightened efficacy in simulating permafrost environments， particularly evident in its more accurate representation of soil temperature dynamics. The mean biases in the simulation of soil temperature and soil moisture at permafrost and seasonally frozen soil stations were determined to be 0.56/-1.40 ℃， and -0.001/0.03 m³·m⁻³， respectively. Further examination of simulations pertaining to active layer thickness and the duration of ground surface freezing revealed relatively higher errors at permafrost station. Conversely， errors were comparatively minimal at permafrost station， evidenced by an average depth error of the active layer thickness of merely 10.6 cm and an average error rate of 3%， affirming the SHAW model’s high precision in permafrost station simulations. Conversely， seasonally frozen soil station demonstrated superior performance in simulating maximum frost depth， with a mean absolute error of 21.0 cm and an error rate of 12%. The overall error in surface freezing duration simulations at permafrost stations （9%） marginally exceeded that at seasonally frozen soil stations （8%）. In addition， the freezing and thawing processes are significantly different among different frozen soil stations. The average freezing speed （6.75 cm·d^-1） of the permafrost station during the simulation period is greater than that of the seasonally frozen soil station （1.33 cm·d^-1）， while the average melting speed （2.11 cm·d^-1） is smaller than the seasonally frozen soil station （3.15 cm·d^-1）. Since the underlying permafrost layer of Dashalong Station serve as an “underground cold source”， the seasonal fluctuation range of soil temperature in deep layers （below 80 cm） is smaller than that of seasonally frozen soil stations. The research conclusion can provide a reference for the study of the freeze-thaw differences between permafrost and seasonally frozen soil in the upper reaches of the Heihe River.

In order to reveal the influence of the initial stress state on the dynamic characteristics of frozen clay under the condition of principal stress shaft rotation， a circular stress path of pure principal stress rotation was realized by using the frozen hollow cylindrical apparatus（FHCA）， and a series of hollow torsional shear tests in laboratory were carried out on the basis of the circular stress path， and the effects of the initial stress state on the cumulative plastic strain， stress-strain hysteresis loop and dynamic modulus of frozen clay were studied. The results show that the greater the initial stress of the specimen under the condition of pure principal stress rotation， the faster the cumulative plastic strain development rate of the frozen clay hollow cylindrical specimen， and the greater the final cumulative plastic strain generated. In addition， it is found that with the increase of the initial stress under the pure principal stress rotation condition， the inclination of the axial stress-strain and shear stress-strain hysteresis curves of the frozen clay specimen increases， and the axial resilience modulus and shear modulus of the frozen clay specimen also increase， and there is a linear relationship between the axial resilience modulus and shear modulus and the static deviation stress ratio under different cyclic stress ratios. This research results are helpful to improve the design theory of cold region engineering and artificial freezing engineering.

The warming and humidifying climate on the Qinghai-Xizang Plateau of China is causing permafrost degradation at a faster rate， which is affecting the stability of infrastructure such as highways and railways on it. To resolve this problem， thermosyphons have been widely applied to protect the degrading permafrost in Northeast China and the Qinghai-Xizang Plateau. The thermosyphon has the characteristic of active unidirectional cooling. Existing researches often use numerical simulation or borehole temperature measurement methods to analyze its cooling effect on permafrost foundations. With the development of geophysical exploration technology， ground penetrating radar （GPR） technology provides a new research method for frozen soil engineering， which can obtain continuous data without causing damage to structures. The section of the Qinghai-Xizang Highway from Tuotuo River to Tanggula Mountain Pass was selected as the research area， and GPR technology is used to study the cooling effect of thermosyphons. Based on the investigation of road problems， three typical road sections were selected， where double-sided and single-row vertical thermosyphons， single-sided and single row oblique thermosyphons， and adjacent non-thermosyphons were placed， respectivley. GPR technology was used to detect and analyze structural damage and underlying permafrost distribution. Meanwhile， combined with on-site investigations， the impact of different thermosyphon placement methods on the cooling effect of frozen soil embankment was evaluated. The results indicate that the cooling effect is as follows： double-sided and single row vertical thermosyphons have better cooling effect， followed by single-sided and single row oblique thermosyphons and non-thermosyphon. The distribution of the permafrost layer beneath the double-sided and single-row vertical thermosyphon embankment had good continuity， with an increase of 0.47 m in the upper limit of permafrost compared to the natural surface near the road， and the road structure was entire. The permafrost layer under the single-side and single-row inclined thermosyphon embankment had general continuity， and its permafrost table was close to that of the natural surface. There were areas where structures loosened and cracks happened in the embankment structure. The continuity of the permafrost layer under the non-thermosyphon embankment was lower， and the degradation of permafrost was significant， its permafrost table has degraded by 0.80 m compared to the natural surface. The embankment structure showed large areas of looseness and cracks， and there was water accumulation in some areas. Meanwhile， permafrost degradation can trigger road engineering damages. Comparison showed that the single-sided and single-row oblique thermosyphon embankment and the adjacent non-thermosyphon embankment with permafrost degradation showed loose embankment structures and developed some cracks， and were prone to uneven settlement， cracks， potholes， and other problems. GPR testing can effectively present the distribution of permafrost beneath roads and the characteristics of embankment structural damage. Furthermore， it can used to analyze the mechanism of road surface problems and to provide scientific basis for highway maintenance. Analysis shows thermosyphons can effectively slow down the degradation rate of permafrost， but different placement methods have a certain impact on the cooling effect， the single-sided layout of thermosyphon for high embankments does not cool the warming permafrost. Therefore， it is necessary to scientifically and reasonably carry out the correct design， standardized construction， and effective operation and maintenance of thermosyphon embankments based on the thermal state and structural characteristics of the roads. This can raise the permafrost table under the embankment and reduce or slow down engineering problems caused by permafrost degradation. This study has important practical significance for promoting thermosyphon to be widely used and for improving the serviceability of highways in permafrost regions.

The accurate simulation and projection of hydrological processes are of great significance for the management and protection of water resources in the Qinghai-Xizang Plateau. However， the limited and uneven distribution of national meteorological observation stations brings a great challenge to the accurate simulation of hydrological processes in plateau areas. Utilizing multi-source precipitation data is an effective means to improve simulation accuracy. In this study， the Golmud River basin in the northern Qinghai-Xizang Plateau was selected as a case study area. The VIC-CAS model was employed using three precipitation products （ERA5， WorldClim， and TPHiPr） to represent the multi-year monthly average precipitation as the background field. The meteorological forcing data was derived through interpolation of observed data from meteorological stations， enabling simulation and analysis of the runoff process in the Golmud River basin. Additionally， future runoff predictions were made by downscaling CMIP6 climate model data. The results showed that the meteorological forcing data obtained by using the annual average precipitation of ERA5 as covariate and TPS interpolation method has the most effective simulation results， The Nash efficiency coefficient （NSE） were 0.71 and 0.70 in the calibration and validation period. The contribution of glacier meltwater and snow meltwater to annual runoff in Golmud River basin is about 14.9% and 32.5% respectively. From 1971 to 2019， under the background of slow increase of annual precipitation and significant trend of temperature rise， the increase rates of snowmelt runoff and glacier runoff were 0.28×10⁸ m³⋅（10a）^-1 and 0.03×10⁸ m³⋅（10a）^-1， respectively. The annual runoff increased at a rate of 0.54×10⁸ m³⋅（10a）^-1， and the increase of snowmelt runoff contributed more than 50%. Under the SSP2-4.5 and SSP5-8.5 scenarios， the average annual runoff of Golmud River basin from 2025 to 2100 is projected to be 10.92×10⁸ m³ and 11.51×10⁸ m³， with an increase rate of 0.38×10⁸ m³⋅（10a）^-1 and 0.51×10⁸ m³⋅（10a）^-1， respectively. Furthermore， it is projected that snowmelt runoff and glacier runoff in the future period will reach their inflection points in the 2020s and 2030s under both SSP2-4.5 and SSP5-8.5 scenarios， followed by a significant decrease by the end of the 21st century. This study provides valuable insights for simulating runoff in other data-scarce regions of the Qinghai-Xizang Plateau.

The changes in Qilian Mountain glaciers are not only a direct reflection of climate change， but also has an important impact on fresh water resources. In recent years， the glaciers in Qilian Mountain have been in a state of retreat and instability， to deeply understand the response of the glaciers in the Qilian Mountain to climate change in the 21st century and the magnitude of the change， clarify the characteristics of the Qilian Mountain glaciers change， the process and the reasons， and further predict the trend and magnitude of glacier change， and analyze the Qilian Mountain glaciers destabilization and imbalance in the climate warming background of the mechanism， we selected the Qiyi Glacier， which is strongly influenced by the westerllies， and the Lenglongling No.2 Glacier， which is strongly influenced by the monsoon as examples by analyzing remote sensing images and numerical simulation， it is found that the two glaciers show obvious retreat at the end of the glacier in 1990 to 2022， with the area reduced by 10.28% and 19.04% respectively， and the retreat rate of Lenglongling No.2 Glacier is obviously larger than that of Qiyi Glacier， which is attributed to the fact that the eastern part of the Qilian Mountain is warmer than the western part， and the eastern part of the glacier has decreased in precipitation. To explore the future changes of the two glaciers， the OGGM model considering glacier flow was driven by CMIP6 model data.The relative errors were within 2.5% by comparing the glacier area of the two glaciers simulated by the model and the visually interpreted area from 2008 to 2022. The simulated values of the mass balance of the Qiyi Glacier were more in line with the observed values， with an average difference of 88 mm w.e. Between the two mass balances， the difference between the two is 88 mm w.e.， which indicates that the OGGM model can simulate the glacier changes better.The simulation results show that under the high-emission scenario （SSP5-8.5）， the area and volume of the Qiyi Glacier will decrease by 77% and 94%， and the area and volume of the Lenglongling No.2 Glacier will decrease by 93% and 99% by 2060， and by the 21st century 80s， the air temperature in the area of the Qiyi Glacier will increase by 4.1 ℃ compared with that of the period of 1990 to 2015， while the increase in precipitation will not be obvious， and the glacier will be completely extinguished by that time. In contrast， temperatures in the eastern section of the Qilian Mountain will rise even faster， and it is predicted that the Lenglongling No.2 Glacier will be completely extinct around 2075. Even under the most optimal carbon emission scenario （SSP1-2.6）， a temperature increase of 1.0 ℃ in the Qilian Mountain by 2050 compared with the 1990 to 2015 prognosis will lead to the retreat of both glaciers into ice buckets by 2080 when the area and volume of Qiyi Glacier will be only 22% and 8% of that of 2020， and the Lenglongling No.2 Glacier will be only 17% and 6%. The rapid ablation of glaciers will cause changes in glacier runoff， and the runoff of the Qiyi Glacier and the Lenglongling No.2 Glacier will peak in 2035 and 2024， respectively， and the runoff of the two glaciers will be reduced by 28.98% and 41.82% by the end of the century under the SSP5-8.5 scenario， respectively. Therefore， regardless of climate scenarios， Qilian Mountain glaciers will retreat significantly in the 21st century， or even disappear， due the temperature， precipitation atmospheric circulation， and other influences， which will also make the eastern Qilian Mountain glaciers than the western glaciers retreat faster. Clarifying the trend and magnitude of glacier changes， and understanding the process mechanism of Qilian Mountain glaciers imbalance and destabilization in the context of climate change will deepen our understanding of the current and future glacier changes in the Qilian Mountain region， so as to cope with the environmental and water resource problems caused by the glacier changes， which require us to plan in advance.

This study is based on 600 meteorological stations distributed in seasonally frozen ground regions of China， the annual maximum freezing depth and temperature observation data were measured in the past 50 years from 1971 to 2020. In the 600 stations， the annual variation trends of temperature and freezing index were discussed respectively. On the basis of Stefan improved formula， the statistical relationship between the E_a factor （based on temperature freezing index） and factors such as longitude， latitude and altitude were further explored on the standard field， and established an empirical formula. Using the empirical formula to calculate the E_a factor， and verify the accuracy of it. The research shows that the E_a between the empirical formula and calculated by Stefan improved formula （based on temperature freezing index） have good fitting degree at most points， but some points at the boundary of permafrost regions have large fitting differences. In order to solve this problem， the research region was redivided into the general seasonally frozen regions and the transition regions at the boundary between the seasonally frozen regions and the permafrost regions. Correlation analysis was carried out respectively， and the relationship between the E_a and factors such as longitude， latitude， altitude were established. By comparing the observed values， it can be found that fitting degree of these points have been significantly improved. The correlation relationship was introduced into Stefan improved formula （based on temperature freezing index） for further expansion， and finally getting the empirical formula based on temperature freezing’s Stefan improved formula， which was used to calculate the freezing depth of the standard field in seasonally frozen ground regions of China. Since the E factor is relatively complicated to obtain， it is not easy to apply to engineering designs and constructions. For engineering construction sites where it is difficult to obtain the freezing depth， it calculated by using the empirical formula is similar to the observed values of meteorological stations. The combination degree is highly， which can be better provide the freezing depth as a reference for engineering constructions.