Ice cores are an effective proxy for past climate changes， and high-resolution reconstruction necessitates precise dating methods. This study combines manual dating with an automated layer recognition algorithm （StratiCounter） based on a Hidden Markov Model （HMM） framework to test， validate， and analyse the performance of the algorithm in assisting the dating of Zuoqiupu， Dasuopu， and Anemaqen ice cores. The method reduces the time required for manual layer-by-layer identification during binding reference layers， synchronizing with other ice core records， and reviewing processes by automatically inferring layer positions and providing error ranges. Initially， a collection of fundamental dating templates and layer constraint points are manually provided. Next， the algorithm familiarises itself with the template for calculating the quasi periodic cycle of the ice core proxy variable. It then identifies the target paragraph and examines the presence of layer constraints. Perform a comprehensive search for potential points while considering layer constraints， and employ the maximum likelihood method to determine the most suitable solution from the available options. The findings show that for the upper sections of the Zuoqiupu and Dasuopu ice cores， the algorithm-assisted dating results have an accumulated error of less than 2 years （3%） when compared to known chronologies， indicating a high level of consistency. Even under simulated sample loss conditions and the absence of tie points， the algorithm remains operational， resulting in a cumulative error of less than five years. Further analysis of the upper section of the Anemaqen ice core reveals 1 to 2 year differences between algorithmic and manual dating results at specific sections e1 ［2.07， 4.47］ and e2 ［8.31， 9.71］. After using pollution layer data， the ice core at 56.48 metres was dated to 74 years with an error range of ±3 years. Integrating reference layer findings like β-activity and 137Cs with δ18O counting layers helped determine the age sequence of the upper 0.05~56.48 m of the Anemaqen Ice Core， from 1947 to 2020 AD. The method combines the advantages of manual dating with the standardised hierarchical recognition capability of algorithms， thereby reducing the need for manual intervention in the age correction process. It also quantifies potential dating errors and， to some extent， improves the dating procedure.
The Pamir Plateau is the largest center of glaciation in the high mountain regions of Asia. Glacial melt water plays a key role in regional water resources and water cycle. However， in recent years， there are still controversies about the understanding of glacier changes in the Pamir Mountains. On the one hand， it is believed that there is an abnormal progression of glaciers in the Karakoram Pamir Mountains Plateau. On the other hand， it is believed that the glaciers in the Pamir Plateau are accelerating their retreat. The reasons for this are twofold： firstly， there is a mismatch in the spatiotemporal scope of the study， and secondly， there are differences in the definition of glaciers in the study. Based on this， in order to understand the trend and magnitude of glacier changes in the Pamir Mountains， this study， based on multi-source remote sensing images and the Google Earth Engine （GEE） platform， constructed a glacier discrimination method combining the multi temporal minimum NDSI and the classification algorithm of glacial surface moraine area， automatically extracted the area change sequence of debris-free glaciers in the Pamir Mountains from 1990 to 2020 without surface moraine coverage， and analyzed the spatio-temporal change characteristics and trends of debris-free glaciers in the region， And clarify whether there is a phenomenon of “Pamir Karakoram Glacier Anomaly”. This study provides a profound explanation of the current changes in glaciers in the Pamir region， which can effectively support regional water resources and ecological environment protection. In order to clarify the spatiotemporal changes of debris-free glaciers in the Pamir Mountains in the past three decades， this study， based on the GEE platform， combined image cloud cover， seasonal snow cover， moraine cover at the end of the glacier and other factors， obtained the spatiotemporal changes of high-resolution debris-free glaciers in the Pamir Mountains in the past 30 years. The results indicate that the combination of GEE cloud computing platform and glacier classification algorithm is effective and feasible for quickly extracting glaciers， improving the efficiency of long-term glacier change research. During 1990—2020， the area of debris-free glaciers on the Pamir Mountains showed a general shrinking trend. In the past 30 years， the area of debris-free glaciers has shrunk from （12 108.98±250.38） km2 to （8 616.44±7.22） km2， with an average reduction rate of 116.42 km2·a-1.The area of naked glaciers in the west of Pamir Mountains Plateau generally shows a shrinking trend， while the area of debris-free glaciers in the east of Pamirs is relatively stable， the shrinking rate is relatively small， and some regions show an increasing trend during 2000—2010. There is an abnormal phenomenon of Pamirs glaciers. Since 2010， the area of glaciers began to decrease， and the abnormal phenomenon of Pamirs glaciers no longer exists. Based on the analysis of DEM data in Pamir Mountains， it is found that the distribution of debris-free glacier area in Pamir Mountains is normal with altitude， and the largest debris-free glacier area occurs at 4 900~5 200 m altitude. Since 1990， the reduction in debris-free glacier area has mainly occurred in mid altitude areas of 4 600~5 500 m， and has decreased towards high and low altitude sides. Based on the analysis of climate and glacier elevation changes， it is concluded that the glacier anomaly on the Pamir Mountains from 1990 to 2010 is caused by the thinning of the glacier thickness and the glacier jumping at the end of the glacier. This article explores the combination and application of high-performance algorithms and large-scale remote sensing data resources from cloud computing platforms with traditional band combination algorithms， enabling GEE to play a more important role in large-scale research such as the entire high Asia region to global glacier changes， and providing new ideas for large-scale spatial scale research and efficient data processing.
The change of lake ice is a sensitive indicator of regional and global climate change. Most of the existing studies have focused on the phenology of lake ice， or the type of surface freezing， while less on its vertical type （i.e. floating ice and bedfast ice）. Compared with floating ice， bedfast ice can provide a protective environment below 0 °C to the lake bed during the entire freezing period， and delay the development and carbon emissions of the talik. Therefore， accurate identification and classification of both ice types are of great significance to many research fields of the cryosphere. Due to the fact that most lakes that have a large amount of bedfast ices are located in remote and cold regions， it is difficult to conduct large-scale and long-term field observations， resulting in limited knowledge of the distribution of bedfast lake ices. Satellite-based synthetic aperture radar （SAR） provides an advanced means for large-scale identification of lake ice types and periodic monitoring their transitions. Since the 1990s， SAR remote sensing has been widely used for identifying bedfast ice， achieving a series of methods and datasets， providing important and advanced technical means for observing bedfast ice in large and remote lakes. In recent years， with the further development of remote sensing technology， SAR in particular， the variety of remote sensing data has become increasingly diverse， and the spatiotemporal resolution has significantly improved， bringing new opportunities and challenges to lake ice studies. At present， most remote sensing studies on lake ice focus on the phenology or thickness of lake ice， but limited attention has been paid to the vertical development of lake ice （whether it can touch the bottom or not）. Up until present， there is no literature review on the use of SAR remote sensing to identify bedfast ice. Therefore， it is necessary to comprehensively and systematically review the research progress in this field， summarize the achievements， identify the existing problems and challenges， and further promote the application of SAR data in bedfast lake ice monitoring， and support related climate change and lake ecology studies. In this paper， a comprehensive review of the research on the classification of bedfast ice and floating ice by SAR remote sensing is conducted. The basic principles and common methods are summarized， and the supporting role and indicative significance of lake ice type for the related research are combed. The results show that： （1） The enhanced backscatter of floating ice is mainly caused by the combined effect of bubbles and high roughness of the ice-water interface， but the backscatter increment produced by each of them at different wavelengths still needs further study； （2） Classification methods have developed from threshold method to object-oriented segmentation， time series clustering and machine learning. However， the accuracy is still limited by factors such as ice melt， snow cover and lake boundary； （3） Lake ice type can reflect the vertical development process of lake ice during the freezing period， therefore it reflects the response of ice thickness to climate change and its impact on the permafrost layer at the lake bottom with different bathymetries， which has indicative significance for the studies of climate， ecology and freshwater resources. Compared to lake ice phenology that focuses on reflecting climate change through the timing of freezing and thawing， bedfast lake ice reflects more about the vertical development process of lake ice during the freezing period. Therefore， it focuses on reflecting the response of ice thickness to climate change and its impact on the permafrost layer at the bottom of different depths of lakes， which is of great significance for climate， ecology， and freshwater resource research. This review summarizes the research progress of SAR remote sensing in identifying bedfast ice， which is anticipated to promote the usage of SAR data （e.g. Sentinel-1） in this field.
In recent years， typical landslide-mudflow disaster chain events induced by strong earthquakes and extreme weather in loess areas have become more frequent. Compared with single disasters， chain geological disasters have stronger concealment， wider spread， higher damage degree and more serious losses. Therefore， the research on the prevention and control of chain geological disasters has always been a hot and difficult point in disaster prevention and reduction. At 23：59 on December 18， 2023， an Ms 6.2 magnitude earthquake struck Jishishan County， Linxia Prefecture， Gansu province， at a depth of 10 km， killing 151 people. The strong earthquake triggered a landslide-mudflow chain disaster in Caotan Village and Jintian Village， Zhongchuan Town， Minhe County， Qinghai Province， resulting in a total of 20 deaths， the disaster chain fatality rate accounted for 13.5% of the total number of earthquake fatalities. After the disaster， a comprehensive and systematic study on the landslide-mudflow disaster chain was carried out by means of remote sensing image processing and interpretation， literature sorting and screening， UAV photographic aerial survey， detailed field survey and visit， and on-site sampling and analysis testing， so as to recover and reappear the start-slide-flow-accumulation process of the landslide-mudflow disaster chain. This paper discusses and puts forward nine major coupling and disaster-causing effects such as surface freezing blocking water retention in the slip source area， gully leveling backfill water retention in the slip source area， platform irrigation seepage filling effect in the slip source area， vibration liquefaction screening effect in the slip source area， growth rate of the funnel closure in the slip source area， soil differential flow sorting effect in the flow area， collapse recharge effect in the flow area， acceleration effect of ice filling water at the bottom of the flow area， and upstream seepage recharge and range extension effect in the accumulation area. The Zhongchuan landslide-mudflow chain disaster caused by the Ms 6.2 earthquake in Jishishan， Gansu Province is a typical secondary geological disaster of the same earthquake， which is transformed into a disaster by the coupling and superposition of multiple disasters such as earthquake， landslide and mudflow， and its impact and effect are much greater than that of a single disaster， and it has the comprehensive characteristics of suddenness， high speed， concealment， confusion， destruction and remoteness. The Zhongchuan coseismic landslide-mudflow chain disaster with nine linkage coupling disaster-causing effects is obviously different from the traditional geological disasters such as landslides and debris flows， and the transformation of disasters and disaster-causing factors are complex and unique. In the process of promoting the construction of beautiful villages and the Belt and Road， the relevant land use development and engineering construction planning must fully consider the risks of regional landslides， debris flows， mountain floods and their coupling disasters， scientifically and rationally lay out land and engineering planning and design， ensure the flooding and silting of valleys， and reserve safe transition areas， so as to prevent disasters and avoid damage to the greatest extent. Long-term and large-scale agricultural irrigation is the most important source of recharge for the loess layer with large thickness and high water content to saturation in the gentle slope area of the high terraces of the Yellow River， and chain disasters occur under the action of strong earthquakes. Therefore， under the condition of ensuring water for agricultural production， strengthening the drainage of the slope zone in the irrigation area， optimizing the irrigation method， and lowering the groundwater level are effective measures to prevent the occurrence of similar landslides and chain disasters and ensure the safety of the geological environment in the loess gentle slope plateau area. The research can provide a scientific reference for the formation mechanism and prevention and control of chain geological hazards.
Noijin Kangsang Region is one of the primary distribution areas of extremely high-altitude glaciers in the mountainous region of southern Xizang （Tibet）. It is situated adjacent to the 349 national highway from Lhasa to Shigatse. The status and changes of glaciers in this Region are of wide concern. Assessing the glacier changes in the region is of great significance for guiding local glacier development， as well as for monitoring and protection. Based on the remote sensing images of the historical period， the glacier boundary of Noijin Kangsang Region from 1976 to 2022 was extracted by using the band ratio and visual interpretation method. The characteristics of glacier area， surface velocity and thickness change in Noijin Kangsang Region were analyzed as a whole， and three typical glaciers in the region were selected for detailed analysis. Combined with factors such as climate， topography and glacier surface albedo， the reasons for glacier changes are explained. The results show that： From 1976 to 2022， the glacier area in Noijin Kangsang Region decreased by （17.88±6.75） km2， it accounted for （17.63±6.70）% of the glacier area in 1976. There are significant differences in the number and area changes of glaciers of different sizes， and the topographic characteristics of glaciers also lead to the heterogeneity of glacier changes. From 2000 to 2019， the average thinning rate of glaciers in Noijin Kangsang Region was 0.26 m·a-1. Glacier thinning was most prominent during 2000—2004. In recent years， the thinning amount and thinning range of regional glaciers have shown a downward trend. From 1988 to 2018， about 62% of the glacier-covered areas in Noijin Kangsang Region showed a deceleration trend. Within the region， eight glaciers have shown significant acceleration， with topography being the primary driving factor for glacier acceleration， while the increase in liquid precipitation at high altitudes may be a major contributing cause to the acceleration. From 1968 to 2022， the three typical glaciers all showed a state of retreat， but the retreat， thinning and surface velocity changes of typical glaciers in different periods had distinctive characteristics. The warming of Noijin Kangsang Region in the past 20~30 years is the main reason for the retreat of glaciers， and the topography affects the retreat， thinning and surface velocity changes of regional glaciers. Simultaneously， the reduction in glacier surface albedo is also one of the reasons driving glacier melting.
Iron （Fe） is an essential nutrient element for biological activities， which can promote the formation of chlorophyll a， and is an important limiting factor for plankton growth in high nutrient salt and low chlorophyll sea area. With the accelerated ablation of glaciers， the erosion of melt water and the abrasion of ice and rock are enhanced， and the rate of chemical weathering is increased， which promotes more Fe elements to transfer from bedrock to melt water. Glaciers transport a large amount of bioactive Fe downstream， which will affect the primary productivity of aquatic ecosystems， thereby affecting the carbon cycle and feedback the evolution of the climate system. The biogeochemical cycle of Fe in glacial meltwater is a hot topic in cryospheric hydrochemistry. At present， some progress has been made in the study of Fe in glacial meltwater， but there are still great deficiencies in the time and spatial scale. The spatial variation and transport process of Fe in meltwater are still unclear. By reviewing the research results of Fe concentration and flux and its stable isotope ratio （δ56Fe） in glacial meltwater， the concentrations of soluble iron （sFe）， colloidal/nanoparticle iron （cnFe）， dissolved iron （dFe） and particulate iron （pFe） in glacial meltwater and their spatiotemporal variations of δ56Fe were analyzed at watershed， regional and global scales. The release rates of various forms of iron from regional and global glaciers were quantified. The annual dFe flux of the glacier is about （403±737） Gg and the production is about （7 151±13 061） kg·km-2. Concentration of Fe in the meltwater is seasonal， and is affected by factors such as bedrock lithology and runoff. The amount of iron output from glaciers is mainly affected by the flow rate. In the future， we should strengthen the research on the temporal and spatial evolution of glacier meltwater iron， analyze its source and transport law by using δ56Fe， pay attention to the bioavailable iron， and comprehensively understand the Fe cycle process and its ecological and climatic effects in the glacier environment.
Optical and/or SAR remote sensing images are frequently used to obtain glacier velocity fields through the use of offset tracking due to its strong decorrelation. Correlation algorithms extract the pixel-level offset， which can then be refined to a sub-pixel level through various interpolation techniques. However， the accuracy of these interpolation algorithms incorporated in different offset tracking software has rarely been assessed or compared. The lack of in-situ observations to confirm the sub-pixel precision of derived offset led to the aim of this study， which is to use a digital image processing method to evaluate the precision for various software and algorithms. Furthermore， the study suggests an algorithm to correct the possible offset tracking bias at the sub-pixel level. This study uses six pairs of Sentinel-2 images that observed two of the largest glaciers in Greenland， Petermann Glacier and Kangerlussuaq Glacier. These two glaciers account for roughly 4% each of the entire ice sheet’s glacier mass loss and flow in northwestern and southeastern directions， respectively. The study combines the offset tracking results obtained from different algorithms， including COSI-Corr， autoRIFT， and ImGRAFT （CCF-O and NCC）， and treats them as pre-set offset fields. Using the Sinc interpolation， which is an optimal interpolation method according to the sampling theory， simulated offset images are generated using the pre-set offset fields and pre-event images. Performing the mentioned software and algorithms， offset tracking results are obtained based on the pre-event images and simulated offset images. As all the algorithms first establish an offset value at the pixel level and then interpolate to the sub-pixel level， presuming the former being more dependable， this research assesses the precision and inspects possible bias at the sub-pixel level only. The displacement results obtained and the pre-set offset fields are wrapped to a range of ［-0.5， 0.5］ and designated as y and x. A cubic function， y=ax+4（1-a）x3 （where a is the correction parameter）， is chosen for the regression. The precision is exhibited by the fitting’s RMSE， while parameter a indicates the presence of bias； if a equals 1， then no bias exists， but if it's not， there is a bias. Tthe inverse function of the fitting can rectify potential systematic errors at the sub-pixel level. According to the regression results， the sub-pixel systematic error of COSI-Corr is negligible and can be disregarded， whereas autoRIFT and ImGRAFT （CCF-O or NCC） display a certain degree of systematic errors in their offset results. Specifically， the values of a are 1.008， 0.778， 0.915， and 0.886 for COSI-Corr， autoRIFT， ImGRAFT（CCF-O）， and ImGRAFT（NCC）， respectively. In COSI-Corr， Sinc function is used to interpolate the correlation coefficient matrix， while ImGRAFT applies bicubic interpolation regardless of the correlation algorithm being CCF-O or NCC. autoRIFT utilizes a rapid Gaussian pyramid upsampling algorithm for estimating the sub-pixel displacement with a precision of 1/64 pixel. After systematic error correction， the autoRIFT algorithm’s RMSE decreased by an average of 0.0054 pixels in a single direction， resulting in the most significant improvement among all algorithms and an 11% increase in precision. This demonstrates the significance of performing sub-pixel systematic error correction. On the other hand， the RMSE of ImGRAFT （CCF-O） and ImGRAFT （NCC） reduced slightly by an average of 0.0014 and 0.0012 pixels in a single direction， respectively. Whether to apply this correction to ImGRAFT depends on the desired level of precision， as it results in only a 1.5% increase. Furthermore， as no noticeable systematic sub-pixel errors were detected， it is unnecessary to apply this correction to COSI-Corr. The regression results of all software/algorithms are similar across different study sites and/or deformation directions， indicating that sub-pixel systematic error is solely dependent on interpolation algorithm. After the systematic correction， all algorithms show reliable results. COSI-Corr and autoRIFT show higher precision than ImGRAFT， with RMSEs of 0.04~0.14 pixels at Kangerlussuaq. Conversely， ImGRAFT shows slightly lower precision with RMSEs of 0.08~0.10 pixels at Petermann and 0.09~0.13 pixels at Kangerlussuaq. ImGRAFT （CCF-O） shows slightly better precision than ImGRAFT （NCC）. Given the much higher computational requirements of autoRIFT relative to the other algorithms， this study recommends combining autoRIFT with a post-correction step for systematic error.
Glaciers are transformed from snowflakes formed by the condensation of atmospheric water vapor at low temperatures. In general， the transformation of snowflakes to glaciers contains three processes. To begin with， the snowflakes falling on the surface of glaciers are automatically rounded and transformed into firn because a system is more stable when its surface free energy is lower. The subsequent step is firn densification， a lengthy and challenging process that turns firn into glacier ice. During firn densification， firn density increases with depth and time due largely to overburden stress from the accumulation of new snow. Finally， glacier ice flows and progressively converts into glaciers under the force of gravity. Firn densification is a highly complicated process since several physical mechanisms operate simultaneously during densification and dominant mechanisms differ at different stages. Understanding the evolution of density and the physics of firn densification is essential for several applications of glaciology. A firn densification model that can simulate the density evolution of firn is crucial for assessing glacier mass balance accurately via the satellite altimetry method. In this method， the differencing of digital elevation models provides a change in glacier volume， which needs to be converted to a mass change by a density model or assumption. The paleoclimate reconstruction of the ice core requires calculating the age difference between the ice and the air trapped in it. A firn densification model is necessary to determine the age of the ice when the bubbles are close-off， which can be coupled to a firn-air model to calculate the age difference. This paper comprehensively analyzes the research methodology and the latest progress on firn densification in polar ice sheets. Numerous firn densification models have been proposed in recent years. These models are categorized as either empirical models （including semi-empirical models） or physical models. Empirical models are often based upon a steady-state assumption. They are formulated as a function of temperature， accumulation rate （which serves as a proxy for stress）， and several tuning parameters. The temperature sensitivity has been improved by taking the effects of seasonal and interannual temperature variations into account. More recently， the impact of surface meltwater on firn densification rates， including meltwater percolation， retention， and refreezing， has been added into firn densification models. Physical models are built upon physical principles by analyzing the change in grain microstructure and its underlying physical mechanism during firn densification. These models were formulated by microscopic parameters （for example， grain radius， bond radius， and viscosity）. Physical models are currently scarce since the physics of firn densification is not fully understood， and the data needed to develop a purely physical model are still lacking. Up to now， the snow-firn transition is based on the theory describing grain-boundary sliding； the firn-ice transition is based on the theory explaining the pressure sintering of spherical powders. Overall， the study of firn densification in polar ice sheets has made great strides over the past few decades. On a macroscopic scale， the growth and deformation of grains are mainly affected by their microstructure. Both empirical and physical models are unsatisfactory because previous studies on firn densification have not yet fully clarified the microstructure of firn and its connection to the macroscopic process. For the empirical models， the theory of wet firn densification is still incomplete. And the applicability of empirical models is limited due to the neglect of the specific physical mechanisms. In addition， it may fail to generalize the steady-state assumption directly to transient scenarios. The physical models employed today are mainly based on idealized assumptions， which may be speculative and unreliable because the physics of firn densification is still uncertain.
Snow plays a critical role in surface radiation， energy， and water cycles and is often used as an important indicator for global change assessment and monitoring. In northern pastoral areas of China， the winter snow depth and snow cover are extensive， with prolonged snow layer maintenance. Excessive snowfall can have disastrous impacts on livestock production， while the meltwater from snow contributes to the improvement of the ecological and hydrological conditions. On the other hand， the speed of snowmelt is a crucial indicator for assessing the occurrence of snowmelt-induced floods. Therefore， snow accumulation and its melting process， along with the influencing factors， are not only focal points for various industries related to ecological and socioeconomic development， but they also have direct implications for disaster prevention and pattern recognition. Since 2020， meteorological departments in Inner Mongolia have successively established and applied a batch of field automatic snow depth observation instruments， enabling the continuous hourly observation of snow depth at multiple forest land stations. This provides a valuable opportunity for analyzing the dynamic changes of snowmelt in forest land. Based on the hourly data from field automatic snow depth observation stations in Hulun Buir forest region of Inner Mongolia， this study， through the analysis of multiple locations and multiple snowmelt processes， as well as the results of regional common feature analysis， explores high-frequency dynamic snowmelt processes and their influencing factors. This research serves as a direct basis for a more accurate understanding of the dynamic patterns of snowmelt in northern forest land and dynamic predictions. The results indicate： （1） Winter snow in Hulun Buir forest region can persist for about 102 to 155 days， generally entering the melting period in early March. The snowmelt process typically lasts for 5 to 18 days and can be divided into continuous and rapid melting stages. （2） When the snow depth is less than 3 cm， the ground snow cover rapidly enters the melting stage. If the average temperature between 10：00 and 20：00 is above 0 ℃， forest snow cover will completely melt within 36 hours. （3） Snowmelt exhibits a pattern of gradual decline followed by a rapid decline and then a slow decline during the day. Compared to grasslands， the peak snowmelt period is delayed and occurs between 10：00 and 15：00. Thermal conditions are the primary influencing factor for snowmelt in Hulun Buir forest region， and when studying snow cover as the affected factor， it was found that the ground temperature at 14：00 is the dominant factor affecting the snowmelt process and rate. （4） Although temperature and snow surface temperature are closely related to snowmelt， ground temperature emerges as the most sensitive factor influencing snowmelt response during the snowmelt process. Ground temperature at 0 cm， at 14：00 and 18：00， is the key factor determining or affecting the snowmelt process and rate in the northern forest land of China. Additionally， snowmelt is largely determined by thermal conditions.
Based on the temperature and precipitation observation daily data of 99 meteorological stations in Xinjiang from 1961 to 2020， the characteristics of regional strong cold air and cold wave in Xinjiang were analyzed by using correlation analysis and linear trend analysis， as well as the dry and wet conditions of strong cold air and cold wave in Xinjiang were recognized using the precipitation data. The results showed that： （1） The frequency of regional strong cold air and cold wave in Xinjiang showed a decreasing trend， which was mainly manifested in the decrease of the frequency of strong cold air and cold wave in autumn， especially in 2010s. （2） The occurrence frequency of cold wave in Xinjiang was relatively close during the three months of winter. In spring and autumn the cold wave occurred most in November， followed by March and April. The occurrence time of super strong cold wave mainly concentrated in December and January， and the frequency of strong cold wave was the highest in February. （3） During the cold wave in Xinjiang， the minimum temperature increased and the cooling amplitude decreased， respectively， and both the mean and maximum precipitation increased significantly since the 1990s. Meanwhile， the cooling amplitude of the regional cold wave in Xinjiang in winter was about 1 ℃ larger than that in spring and autumn. （4） The minimum temperature and the maximum precipitation during cold wave process occurred in the Altai Mountains and Tianshan Mountains. In spring and autumn the cold wave in Xinjiang was usually normal or dry process， while in winter 76.4% processes were normal dry， and 85.7% of the wet and wet processes were concentrated in the 1990s. The preliminary identification of the dry and wet conditions of the cold wave was helpful to understand the new climate characteristics of the cold wave in Xinjiang more comprehensively and serve for disaster prevention and mitigation. （5） In winter， most stations in Northern Xinjiang and Tianshan Mountains reached the cold wave standard in 72 hours， while in Southern Xinjiang， most stations reached the cold wave standard in 24 hours.
As an important type of land cover， the change of snow cover has important feedback and regulation effects on the local hydrological environment and phenological changes. Based on the daily snow depth observation data of 62 meteorological observation stations in Heilongjiang Province from 1961 to 2020， the temporal and spatial variation characteristics of maximum snow depth in Heilongjiang Province and its relationship with atmospheric circulation， temperature and snowfall factors were analyzed by Mann-Kendall test， empirical orthogonal function （EOF）， correlation analysis and other methods. The results showed that the average maximum snow depth in Heilongjiang Province from 1961 to 2020 was 16 cm， 14 cm， 10 cm and 8 cm in autumn， respectively. The maximum snow depth in year， winter and spring showed a significant increase trend， with an increase rate of 1.40 cm⋅（10a）-1 （P<0.01）， 1.51 cm⋅（10a）-1 （P<0.01）， and 0.76 cm⋅（10a）-1 （P<0.05）， with an insignificant increase trend in autumn. From 1961 to 2020， the maximum annual and seasonal maximum snow depth in Heilongjiang Province changed abruptly in the late 20th century and early 21st century， and the maximum snow depth showed an increase in annual variation after the mutation. The maximum snow depth in Heilongjiang Province showed the spatial distribution characteristics of mountains （Greater and Lesser Khingan Mountains， Wanda Mountains） than plains （Songnen Plain and Sanjiang Plain）， while the change rate was that the plain was larger than the mountainous area， among which the maximum snow depth in the Songnen Plain increased the most. There are two main changes in the maximum snow depth in Heilongjiang Province： east-west reverse type and southeast-northwest reverse type. Temperature， snowfall， polar vortex intensity in the northern hemisphere， and East Asian trough intensity all affected the maximum winter depth in Heilongjiang Province， and the influence of snowfall and polar vortex intensity in the northern hemisphere was greater than that of temperature and East Asian trough intensity. As the climate warms， the influence of temperature and the intensity of polar vortexes in the northern hemisphere on the maximum snow depth in winter becomes more pronounced.
Based on the daily air temperature data of 519 stations in China from 1956 to 2017， this paper displays two main regional spatial modes of winter surface air temperature anomalies in eastern China through the rotating empirical orthogonal function analysis method， and studies the relationship between them and atmospheric circulation. The results show that these two regional spatial modes are relatively stable. The first mode is manifested as the anomalous low temperature in most parts of southern China， which is called southern mode. The extreme cold centers of this mode are located in the south and southeast coastal areas. The anomalous temperature signals of the second mode are mainly concentrated in the northeast China， which is called the northeast mode. There is a significant negative correlation of the time series of southern mode with sea surface temperature in the equatorial central Pacific， and a significant positive correlation with the opposite variation mode of East Asian winter monsoon in the south and north， respectively. When the sea surface temperature in the equatorial central Pacific is anomalously cold in winter， the southern part of East Asia shows a strong low-level northerly wind causing anomalously low temperature in most parts of southern China. At this time there is no significant northerly wind anomaly in the northeastern Asia. By contrast， the winter Arctic Oscillation is significantly related to the northeast low-temperature mode. When the Arctic Oscillation is in its negative phase， the cold high pressure is mainly concentrated in the polar region， and the northeasterly wind prevails on its east side， leading to anomalously low winter air temperatures in Northeast China.
The stability of permafrost roadbeds is crucial for the safe operation of highways and railroads in cold regions. In order to comprehensively understand the current research status and development trend in the field of frozen soil roadbeds， CiteSpace was used to visualize and analyze the literature related to permafrost and seasonally frozen subgrade in the China National Knowledge Infrastructure （CNKI） and Web of Science databases. CiteSpace is a citation space metrics analysis tool that can help clarify the research trajectory， hot topics， and direction of development in a particular research area. Our search time range was from January 1， 2000 to August 31， 2023， and a total of 1 027 Chinese and 854 English documents were included. The current status and trend of research on permafrost roadbeds were illustrated by interpreting and analyzing the mapping of the number of publications， the mapping of authors’ cooperation relationship， the mapping of institutions’ cooperation relationship， the mapping of keyword clustering， and the mapping of keyword clustering timeline in the past five years. The results show that： since 2000， the publication volume of permafrost roadbed research has been on an overall growth trend. Starting from 2017， the number of English literatures gradually exceeded Chinese literatures， of which the publications of Chinese scholars in international journals accounted for about 80%. The cooperation between domestic research institutions is closer. As the core research institutions， the Northwest Institute of Ecology and Environmental Resources of the Chinese Academy of Sciences and Beijing Jiaotong University have formed a good cooperative relationship with other institutions， and have an important position in the journals concerning frozen soil roadbeds both at domestic and abroad. Among the foreign research institutions， Laval University and University of Manitoba in Canada have close cooperation. However， the cooperation between international institutions needs to be strengthened. The keyword clustering mapping shows that the research topics of permafrost roadbeds include three main topics， which are the stability of perennial permafrost roadbeds， the temperature field of permafrost roadbeds， and the diseases and prevention of roadbeds under freeze-thaw cycles. The successful operation of Qinghai-Tibet Railway has fully proved that active cooling measures such as thermosyphon， duct-ventilated embankments and crushed rock embankment can effectively maintain the stability of roadbed in perennial permafrost area. But the cooling effect of individual cooling techniques is limited in warm and ice-rich permafrost zones and wide roadbeds for highways； in these cases， stability requires composite roadbeds or new， more effective roadbed constructions. In the seasonal frozen soil area， the frost heave mechanics and engineering measures of roadbed are the emphasis of research. Three measures are usually used to prevent frost heave： soil replacement method， water proof and drainage， and heat insulation. On researching hot issues， permafrost experts and scholars have focused on the research of water and vapor migration mechanism of roadbed， frost heave prevention and control of high-speed rail roadbed， and new roadbed stabilization measures in permafrost region. New materials and technologies are constantly introduced into the research of frozen soil roadbed. Ground source heat pump technology， solar energy， etc. have been proved to be applied to the stability protection of roadbeds. Environmentally friendly materials such as ionic curing agent， microbial conditioning and fiber reinforced materials can be used as external admixtures to improve the freeze-thaw resistance of soil. In the future， with the integration of new generation technologies such as material science， BIM， artificial intelligence， new energy， interferometry synthetic aperture radar （InSAR） with mechanization， the stability reinforcing measures for permafrost roadbeds and the anti-freeze-thaw measures for seasonal permafrost roadbeds will be more environmentally friendly while improving the effect. The construction， management and monitoring of permafrost roadbeds will also continue to evolve towards digitalization and intelligence.
Recycled concrete refers to the new concrete prepared by crushing and screening construction waste， mixing according to a certain gradation， and replacing natural aggregate， which can effectively alleviate the problem of recycling of construction waste and shortage of natural sand and gravel materials in China. As a porous building material， the transfer of water is an important reason for the deterioration of concrete performance. The migration of water in concrete mainly includes three forms： capillary water absorption， infiltration and diffusion. In unsaturated concrete， capillary water absorption is the main way of water transmission， and it is also an important control factor affecting the frost resistance of concrete. In this paper， effects of the number of freeze-thaw cycles and coarse aggregate replacement rate of recycled concrete （0%， 20%， 40%， and 60%） on the capillary water absorption performance of recycled aggregate concrete （RAC） were studied by ASTM C1585-13 standard test method. The prediction model of initial capillary water absorption of RAC under freeze-thaw cycles was established. Combined with the unsaturated capillary theory， the prediction model of relative water content distribution of RAC under different numbers of freeze-thaw cycles was established. The results show that the cumulative water absorption and initial capillary water absorption of RAC increase with increasing replacement rate of recycled coarse aggregate under the same freeze-thaw cycles. When the replacement rate is the same， the cumulative water absorption and initial capillary water absorption of RAC increase with increasing number of freeze-thaw cycles. When the content of recycled coarse aggregate is 20% and 40%， the capillary water absorption resistance is very similar. From an economic point of view， concrete containing 40% recycled coarse aggregate has better resistance to frost. The capillary water absorption of RAC increases rapidly in the early stage， decreases gradually in the middle stage， and tends to be stable in the later stage. It can be seen from the microscopic experiments that the freeze-thaw cycle gradually accumulates the internal damage of RAC， which in turn significantly increases the initial capillary water absorption. Therefore， the influence of frozen soil environment on the durability of RAC should be fully considered in cold regions. It is verified that the prediction model of initial capillary water absorption of RAC under freeze-thaw cycles based on regression analysis has high calculation accuracy and can be used to predict the capillary water absorption performance of RAC， which provides a theoretical basis for evaluating frost resistance durability.
In order to investigate the shallow sliding mechanism of swelling soil channel slopes in the seasonal permafrost zone， nuclear magnetic resonance （NMR） tests， magnetic resonance imaging （MRI） tests， and indoor direct shear tests were conducted on soil samples with unidirectional freezing mode in comparison with multidirectional freezing conditions. The connection between the changes of water content and porosity and mechanical properties of expansive soils under unidirectional open freezing conditions was investigated. The results show that more accurate water content distribution curves after water migration can be obtained by MRI tests. Under the multidirectional freezing condition， the freeze-thaw action reduces the number of large pores and increases the number of medium pores in the soil， and the pore size generally decreases. In contrast， under one-way open freezing conditions， the special growth mode of the ice lens body leads to an increase in the number of both large and medium pores in the soil， and they are concentrated in the location where the ice lens body is developed. The shear strength of unidirectional frozen soil in situ freezing zone is approximately equal to that of multidirectional frozen soil， and the shear strength of the soil in the sub-consolidation freezing zone is lower than that of the in situ freezing zone， which has a greater influence on the internal friction angle. And combined with the investigation of the landslide in the diversion dry canal of the northern Heilongjiang Province， we get that the decrease of soil shear strength due to the development of micro-thin lamellar structure ice lens during spring thaw is the main reason for the slope instability of the swollen soil channel in the seasonal permafrost region.
In the loess area of northwest China， the subgrade engineering is facing the problem of freeze-thaw diseases owing to the freezing-thawing cycles. A commonly prevention method is using the improved loess as subgrade filling， and now the approach of poor subgrade soil improved by the industrial waste has become a new trend of ground treatment. Considering a huge stock of coal gasification slag and their low utilization rate， and coal gasification slag needs to be immediately treated safely and reasonably. Thus， we proposed a new approach of using the loess improved by the coal gasification slag as subgrade filling in seasonal frozen ground regions， and our previous studies have shown that a certain amount of coal gasification coarse slag improved loess as subgrade filler is suitable for mechanical properties. In this study， the freeze-thaw characteristics and microscopic mechanism of coal gasification coarse slag modified loess under freeze-thaw cycle conditions were explored through laboratory freeze-thaw cycle tests， electron microscope scanning and CT scanning tests， and the two groups of pure loess and single lime modified loess were tested as comparison groups. The experimental results show that： （1） After the first and second freezing-thawing cycles， the frost-heave ratios of the group of mixing 15% coarse coal gasification slag were 0.02% and 0.29%， and the thaw-settlement coefficients were 0 and 0.05%. The frost-heave rates of the groups of pure loess and with 4% lime were 0.43% and 0.63%， 0.38% and 0.42%， respectively. The thaw-settlement coefficients were 0.26% and 0.22%， and 0 and 0.13%， respectively. However， after three times of freeze-thawing cycles， the frost-heaving ratio and the thaw-settlement coefficient of the experimental group with only 15% coarse coal gasification slag began to increase， and the average frost-heaving rate and the thaw-settlement coefficient during five freezing-thawing cycles were 0.38% and 0.17%， respectively. It is close to or even greater than the values of the experimental groups of pure loess and mixing 4% lime， that is， the average frost-heave rate and thaw-settlement coefficient in the first two freezing-thawing cycles decreased by 70.8% and 89.6%， respectively， compared with the pure loess group， indicating that the group of mixing 15% coarse coal gasification slag can significantly restrain the frost heave and thaw settlement of loess at the initial stage of the freezing-thawing cycle action， but with the increase of freezing-thawing cycles numbers， the effect of inhibiting freezing-thawing on loess weakened gradually. （2） The temperature gradient at each position of the sample with 15% coarse coal gasification slag is the smallest， followed by that of the sample with coal gasification coarse slag and lime， and the thermal conductivities of the two groups of samples were the smallest under positive and negative temperature conditions， indicating that the two groups can significantly reduce the temperature fluctuation at each position of the sample， and will reduce the thermal conductivity of the sample. （3） After five freeze-thaw cycles， the water content of the sample with 15% coarse coal gasification slag is the lowest at each height， indicating that the addition of coal gasification coarse slag is conducive to controlling the distribution of water content of the soil sample after the freezing-thawing cycle. （4） The freeze-thaw characteristics of loess can be significantly improved by adding an appropriate amount of coal gasification slag at the initial stage of the freeze-thaw cycle action， and the porosity of the sample before and after the freezing-thawing cycle can be reduced by 41.5% and 47.8%， respectively， compared with that of pure loess， making the soil much denser. In general， the addition of coal gasification slag in loess can improve the service ability of loess subgrade fill in cold regions when the number of freeze-thaw cycles is less than 3， but its freeze-thaw characteristics under more freeze-thaw cycles need to be further studied. This paper puts forward some new thoughts on the stability of loess roadbed engineering and the utilization of coal gasification slag resources in seasonal frozen ground regions.
The dynamic relationship between the unfrozen water and ice in frozen soil at a given negative temperature is closely related to its initial moisture content. The variation of unfrozen water content significantly affects frozen soil microstructure. Herein， giving a quantitative relationship between unfrozen water content and the microstructure of frozen soil under different initial water contents is crucial for revealing the special physical and mechanical properties of frozen soil. In this study， we first investigated the microstructures of frozen clay with different initial water contents （11.8%， 23.1%， 32.2%， 42.2%， 54.1%， 64.2% and 74.8%） under different test temperatures by using Cryo-scanning electron microscope （Cryo-SEM）. Then， the nuclear magnetic resonance （NMR） technology was carried out to monitor the unfrozen water content of frozen soil under different initial water contents and different temperatures. Based on the tested results， a quantitative relationship among the initial water contents， unfrozen water contents， and microstructure of frozen soil under a given negative temperature were given. The results demonstrated that as initial water content increases， the pore structure of frozen clay develops from inter-aggregate pores to inter-granular pores when the water content is below the liquid limit. When the water content exceeds the liquid limit， the pores within between super-aggregates start to develop as the initial water content rises. The pore fractal dimension is taken as a parameter to characterize the degree of confusion of soil pores in this study. It has a large pore fractal dimension， which indicates that the soil pores are more chaotic. The soil with lower pore fractal dimension has more uniform and regular pores， and the better the orientation of soil pores. the pore fractal dimensions are extracted by processing the images of Cryo-SEM by MATLAB software. The study results show， under the same negative temperature condition， the pore fractal dimension and pore area increase after increasing initial water content. For the type of relatively high-temperature frozen soil at -2 ℃， the pore fractal dimension increases with the increase of unfrozen water content along with initial water content increasing， while for the types of relatively low-temperature frozen soil at -5 ℃ and -10 ℃， the pore fractal dimension increases along with unfrozen water content decreasing. Finally， a quantitative relationship was established to describe the microstructure of frozen clay and its unfrozen water content under varying initial water content conditions at a given negative temperature. These researches have an important significance for the sustainable development and construction of engineering in cold regions.
In recent decades， with the global warming and the enhancement of human activities， the trend of permafrost degradation is significant. This process changes the hydrogeological conditions of the cold region， thus causing a series of ecological environment changes. Thermokarst lakes are the product of permafrost degradation and active thermokarst lakes indicate the decrease of permafrost stability. Thermokarst lakes are also a major source of greenhouse gases， which are important in the carbon cycle in cold regions. The formation and development of thermokarst lakes will cause the change of frozen soil environment， which will affect the global climate change. Therefore， in view of the increasingly serious thermokarst phenomenon in the process of permafrost degradation， and the resulting problems such as permafrost thawing penetration， groundwater level change， thermokarst lake expansion and so on， the current research status of the hydrology change of thermokarst lake in permafrost area is analyzed and sorted out. This paper mainly discusses from the following five aspects： （1） Analysis of the main factors of the formation and evolution of the thermokarst lake； （2） Analysis of temporal and spatial variation of thermokarst lake hydrological process and its influencing factors； （3） Thermokarst lake water balance process and influencing factors； （4） Influence of thermokarst lake changes on regional water quality； （5） Influence of thermokarst lake on carbon cycle. Finally， the main problems in the field of thermokarst lakes are discussed： there are some defects in the identification of thermokarst lakes； the factors related to the area change of thermokarst lake are single. There is still a gap in understanding the broader relationship between permafrost gradient， permafrost degradation， and the corresponding hydrological response of thermokarst lakes. Knowledge of the effects of permafrost degradation on lake water chemistry is limited. Based on the existing problems， it is proposed that isotope technology should be used to fully consider the surrounding environment of the pond in the later research， and multi-factor analysis should be carried out in combination with climate change and permafrost degradation， in order to provide reference for further research on the frozen soil hydrological process， regional water resources evolution and the impact on the ecological environment under the background of permafrost degradation.
Climate change （CC） and land use/cover change （LUCC） have been identified as the two main drivers of global hydrological dynamics. Investigating the impact of these factors on watershed hydrological processes is of great significance in unraveling regional hydrological and environmental transformations. In this study， we focus on the Zamu River basin， a representative region in arid Northwest China， and adopt a comprehensive approach utilizing multi-source remote sensing images， hydrometeorological site monitoring， reanalysis product data， the Soil and Water Assessment Tool （SWAT） model， the Mann-Kendall （M-K） trend test， and land use dynamic degree analysis. By integrating these methods， we aim to provide a more accurate and comprehensive understanding of how CC and LUCC affect hydrological processes in the Zamu River basin. By isolating the driving factors of hydrological processes and considering the baseline period， as well as the separate and combined influence of CC and LUCC， we investigate the response characteristics and driving mechanisms of hydrological processes to these factors in the Zamu River basin. The research findings were as follows： （1） Over the period of 1980 to 2014， the annual precipitation， average temperature and runoff in the basin exhibited an upward trend， with increasing rates of 0.47 mm·a-1， 0.067 °C·a-1 and 1.43 mm·a-1， respectively. （2） The distribution of land use types in the basin displays spatial heterogeneity， with grassland， woodland， and unused land being the predominant land uses. The comprehensive land-use dynamic degree was 2.91%， and construction land has the highest single land-use dynamic degree. （3） Aside from the single effect of LUCC， the runoff of the basin decreased by 0.12 mm， while the runoff of the single effect of CC and the combined effect of CC and LUCC increased by 6.27 mm and 6.15 mm， respectively. The evapotranspiration of each scenario showed an increasing trend. CC was identified as the primary driving force for watershed runoff and evapotranspiration. Management strategies for water resources， that enhance the adaptability of the basin to CC， were critical to promote the rational allocation of water resources in the upstream， midstream， and downstream of the basin， in the future.
Real-time and accurate medium- and long-term daily runoff prediction is of great significance for the rational utilization of water resources in arid and semi-arid areas. Aiming at the shortcomings of long short-term memory （LSTM） model such as equal length of input and output time steps， many forgetting when dealing with long sequences， and inability to assign weights according to importance， an LSTM-Seq2seq-Attention based on attention mechanism （Attention） optimization is constructed. The model sets the encoder and decoder in the sequence to sequence （Seq2seq） model as a three-layer LSTM structure， and introduces an attention mechanism to further optimize the model before the decoder outputs the sequence. In order to verify the effectiveness of the LSTM-Seq2seq-Attention model， this study takes the upper reaches of the Dang River as the research area， and simulates and predicts the future daily runoff of the basin in the future 1~7 d based on historical data. The prediction results were compared with those of traditional machine learning models， i.e.， support vector machines （SVM） and the single LSTM model. The results show that the SVM， LSTM and LSTM-Seq2seq-Attention models can be used for short-term daily runoff prediction. However， in contrast， the LSTM-Seq2seq-Attention model has a more prominent predictive effect in medium and long-term daily runoff prediction. This shows that the LSTM-Seq2seq-Attention model has stronger predictive power than a single model， and can be used as a reliable tool for medium and long-term daily runoff prediction and simulation in arid and semi-arid areas.
The Qilian Mountains serve as a significant ecological security barrier of China. The shallow mountainous area of Qilian Mountains is the transition zone that connecting the alpine mountain area and the oasis plain of Hexi corridor. The soil hydro-thermal change is of great importance to maintain the stability of the fragile ecosystem in this region. In this study， the SHAW model was used to simulate the soil temperature and moisture dynamics of the desert grassland in the shallow mountains of Qilian Mountains. The influence of temperature and precipitation variation on soil temperature and moisture in different freezing-thawing stages was quantified by using one-factor-one-time method. The findings indicated that the NSE of the simulated soil temperature in each layer was larger than 0.95， and the simulation performance improved with the increasing in soil depth. The simulated soil moisture could still capture the pattern of soil moisture fluctuation and the rainfall-infiltration process in shallow soil. It should be noted that the distinct freezing-thawing stages result in varied soil temperature and moisture distribution profile features. The soil moisture increased with the increasing in soil depth， while， the variation of soil temperature was gradually diminished. The hydro-thermal process of shallow soil was more sensitive to changes in the surface weather conditions than that in the deep soil. The scenario analysis indicated when the temperature increases by 1.0 ℃， the freezing development stage， the complete freezing stage and thawing development stage was shorten by 1.4 days， 0.8 days and 2.2 days respectively， meanwhile， the temperature in surface and deep layer were increased by 0.6 ℃ and 0.1 ℃ respectively， and the range of the increase in soil temperature was decreased with the increasing in soil depth. The impact of precipitation variation on soil moisture at 0.20~1.20 m is most pronounced during the complete thawing stage， and only responded in the shallow soil at 0.20~0.60 m during the freezing development stage and the thawing development stage， while it is basically not effect on unfrozen water content in the complete freezing stage. These findings can provide reference for maintaining the stability of the fragile ecosystem in the shallow mountainous area of Qilian Mountains under the background of climatic change in the future.
As an important part of the cryosphere， under the background of global warming， this paper studies the freeze-up dates of major rivers in Heilongjiang Province， so as to provide a scientific basis for disaster prevention and mitigation. Based on the observation data of the freeze-up dates of the hydrological observation stations in Heilongjiang Province from 1962 to 2020， this paper used the Mann-Kendall mutation test and linear trend analysis method to explore the characteristics of the river freeze-up dates of Heihe Station of Heilongjiang River， Harbin Station of Songhua River， Jiangqiao Station of Nenjiang River and Raohe Station of Wusuli River with the influence of meteorological factors. The freeze-up date， average temperature， surface temperature， minimum temperature， average wind speed and average sunshine hours were analyzed， and the freeze-up dates of Heihe Station， Harbin Station， Jiangqiao Station and Raohe Station were simulated and predicted by information diffusion theory and multiple linear regression analysis method. The results showed that： （1） From 1962 to 2020， the average freeze-up date of Harbin Station， Jiangqiao Station， Heihe Station and Raohe Station was between November 12 and November 22. The results of Mann-Kendall mutation test showed that from 1962 to 2020， Harbin Station， Jiangqiao Station， Heihe Station and Raohe Station all had mutation dates， and the mutation years were 1970， 2005， 2000 and 2012， respectively. Over the past 59 years， the freeze-up dates of Jiangqiao Station and Heihe Station have been significantly postponed （P<0.05）， with a change rate of 2.46 d·（10a）-1， 1.35 d·（10a）-1， which was postponed by 15 d and 8 d， respectively. （2） The results of correlation analysis show that the average temperature， surface air temperature and minimum temperature are the key factors affecting the freeze-up date of Heihe Station， Harbin Station， Jiangqiao Station and Raohe Station. Specifically， the freeze-up dates of Jiangqiao Station and Heihe Station in relatively high latitude areas are mainly affected by the average temperature， surface temperature and minimum temperature in early November， while the Harbin Station and Raohe Station are mainly affected by the average temperature， surface temperature and minimum temperature in mid-November. （3） According to the information diffusion theory， when the negative accumulated temperature is -180 °C， the probability of river freeze-up at Heihe Station， Harbin Station， Jiangqiao Station and Raohe Station reaches 80%~90%， and when the negative accumulated temperature reaches -240 °C， the rivers of Heihe Station， Harbin Station， Jiangqiao Station and Raohe Station are basically frozen. （4） Based on the key factors such as mean temperature， surface air temperature and minimum temperature， this paper uses multiple linear stepwise regression analysis to simulate and predict the freeze-up dates of Harbin Station， Jiangqiao Station， Heihe Station and Raohe Station. The forecast accuracy of the constructed multiple linear regression model for river freeze-up forecast is more than 80%. It has a good effect on predicting the freeze-up date of Harbin Station， Jiangqiao Station， Heihe Station and Raohe Station， which provides a scientific basis for disaster prevention and mitigation in this area.
The Arctic region is recognized as one of the most climate-sensitive areas， with tundra ecosystems playing a vital role within the Arctic ecosystem. This study employed the ecological niche model named MaxEnt. Utilizing species occurrence data and environmental information， we modeled the current （1970—2000） potential distributions of six dominant tundra species in Alaska and projected their changes under different scenarios （SSP1-2.6， SSP2-4.5， SSP3-7.0， and SSP5-8.5） for the period 2021—2040. The primary factors determining species distribution were identified based on their contribution rates. The results indicate that temperature is the most important environmental factor affecting species distribution， with spatial heterogeneity observed in the changes in suitable habitats for dominant species. Compared to the current climate， the dominant shrub species， Arctous alpina， have shown a decrease in their distribution area. In contrast， the dominant lichen species， Cladina rangiferina， and the dominant sedge species， Eriophorum vaginatum， have exhibited an overall increase in their distribution areas. Under the low radiative forcing scenario （SSP1-2.6）， the suitable habitat areas for Arctous alpina and the creeping dwarf shrub species， such as Dryas integrifolia， have expanded. However， in scenarios with medium， medium-high， and high radiative forcing， which exacerbate global warming， the suitable habitat areas have decreased. Regarding the semi-creeping dwarf shrub species， including Cassiope tetragona， and the dominant moss species， Hylocomium splendens， their suitable habitat areas displayed irregular changes across different climate scenarios. However， it was consistently observed that the low-suitability habitat areas decreased， with a tendency to shift towards moderate to high-suitability regions. Simultaneously， dominant species are shifting towards higher latitudes and elevations.
To investigate the spatial-temporal dynamic characteristic of grassland net primary productivity （NPP） on the Qinghai-Xizang Plateau under global climate change， and provide a scientific basis for ecological environment protection of Qinghai-Xizang Plateau grassland. Taking the Qinghai-Xizang Plateau as the study area， based on the DAYCENT model to simulate the NPP of the Qinghai-Xizang Plateau grassland from 1978 to 2020， the temporal and spatial variation characteristics of the grassland NPP on the Qinghai-Xizang Plateau from 1978 to 2020 were studied. The climate and topography were analyzed using Pearson correlation coefficient and partial correlation coefficient effects of factors on grassland NPP. The findings are as follows： （1） From 1978 to 2020， the NPP of the Qinghai-Xizang Plateau showed an overall upward trend， the total NPP fluctuated from 4.51×105 to 5.77×105 Gg C·a-1， and the annual average was 499.83 g C⋅m-2. In space， there is a decreasing trend from southeast to northwest. （2） The alpine meadow is the main vegetation type of the Qinghai-Xizang Plateau grassland， and its NPP is consistent with the Qinghai-Xizang Plateau grassland in terms of spatial distribution， and the spatial distribution of the alpine meadow shows a decreasing trend from northeast to southwest. （3） Among the effects of climatic factors on grassland NPP， temperature is the main factor causing the increase of grassland NPP. Among the effects of terrain factors on grassland NPP， altitude is the main factor affecting NPP. In the influence of terrain combination factors on NPP， the high-value areas of NPP on the Qinghai-Xizang Plateau are mainly concentrated in the gentle slope areas on the sunny side at low altitudes， and the high-value areas are mainly located in the slope areas on the shady side at high altitudes. In different regions and under different climatic conditions， the NPP values of Qinghai-Xizang Plateau grasslands are significantly different， and the low-altitude sunny side is more conducive to the growth of grassland NPP.
Phosphorus is essential for plant growth and development， existing in various forms in the soil. These different forms of phosphorus vary in their mobility and biological availability. In forest ecosystems， the return of nutrients through litter is crucial for maintaining the dynamic balance of soil phosphorus. With climate warming， changes in the amount of litter in forests are expected， particularly in cold region of China， where significant variations are anticipated. To investigate the effects of litter on soil phosphorus components in Betula platyphylla forests in the permafrost region of the Greater Khingan Mountains， this study employed methods of litter addition and removal to analyze the characteristics of soil phosphorus components following alterations in litter and identify influencing factors. The aim is to provide theoretical references for researching soil phosphorus availability and influencing mechanisms in forest ecosystems in the permafrost region. This study selected Betula platyphylla forests of different ages （30 a， 45 a， and 66 a） in the permafrost region of the Greater Khingan Mountains as the study site. In October 2021， three sample plots measuring 20 m×30 m were established in typical areas of Betula platyphylla forests of different ages. Within each plot， three sample plots measuring 2 m×2 m were designated for litter removal and litter doubling treatments， alongside three control plots with no intervention. In August 2022， soil samples were collected from the 0~5， 5~10， and 10~20 cm soil layers. The Sui’s modified Hedley phosphorus classification method was used to determine the eight phosphorus components. These components were categorized into three groups based on their ease of absorption and utilization by plants： easily-available phosphorus （H2O-Pi， NaHCO3-Pi， and NaHCO3-Po）， moderately-available phosphorus （NaOH-Pi and NaOH-Po）， and stable phosphorus （HCl-Pi， HCl-Po， and Residual-P） to explore the effects of litter addition and removal on soil phosphorus components in Betula platyphylla forests of different ages in the permafrost region. The results showed that litter doubling increases the content of H2O-Pi in different soil layers of the three Betula platyphylla forests， although the magnitude of increase diminishes as forest age increases. Additionally， it elevates the NaHCO3-Pi content， with a significant increase observed in the 10~20 cm soil layer of Betula platyphylla forests of all ages （P<0.05）. However， its impact on NaHCO3-Po content is relatively small. The NaOH-Pi and NaOH-Po content in different soil layers of Betula platyphylla forests of varying ages also showed increases. Specifically， the soil NaOH-Pi content in the 30 a and 45 a Betula platyphylla forests exhibited a significant increase （P<0.05）. Moreover， the NaOH-Po content in the 0~5 cm soil layer of the 45 a and 66 a Betula platyphylla forests showed a significant increase （P<0.05）. On the other hand， litter removal reduced the H2O-Pi and NaHCO3-Pi content in different soil layers of the three Betula platyphylla forests， with a greater impact observed in the surface layers compared to the deeper layers. Additionally， litter removal significantly affected the NaHCO3-Po content only in the 5~10 cm soil layer of the 30 a Betula platyphylla forest， while its impact on other ages and soil layers was relatively small. Furthermore， litter removal reduced the NaOH-Pi and NaOH-Po content in different soil layers of birch forests across all three age groups， with significant reductions observed in the NaOH-Pi content of the surface layers of both the 30 a and 45 a Betula platyphylla forests （P<0.05）. Moreover， litter removal significantly affected the NaOH-Po content in the 5~10 cm soil layer in all age forests. Litter treatment had a relatively minor effect on the soil stable phosphorus （HCl-Pi， HCl-Po， and Residual-P） in different soil layers of the three Betula platyphylla forests. Apart from a significant difference observed in the HCl-Po content of the 0~5 cm soil layer of the 45 a Betula platyphylla forest， no other significant differences were detected （P>0.05）. The redundancy analysis showed that the driving factors of soil phosphorus components were different at different ages. The soil phosphorus components were mainly affected by pH and soluble organic carbon in the 30 a Betula platyphylla forest. In the 45 a Betula platyphylla forest， the main influencing factors are pH and NO3--N. Meanwhile， in the 66 a Betula platyphylla forest， soil moisture content and soluble organic carbon are the primary influencing factors. These findings demonstrated the litter addition and removal exerted a significant impact on the soil easily-available phosphorus and moderately-available phosphorus in three Betula platyphylla forests. However， their influence on stable phosphorus is relatively minor. Increased litter input promotes the activity of phosphorus in the soil of cold-temperate Betula platyphylla forests， enhancing the soil phosphorus availability. Consequently， it is important to safeguard litter in the management of Betula platyphylla forests.
The utilization of wind energy resources is the top priority to achieve the “carbon peaking and carbon neutrality goals”. Based on daily wind speed observation and simulations from 5 CMIP6 models with 7 Shared Socioeconomic Pathways， this study constructs a database with 1 km horizontal resolution and analyses the spatiotemporal changes of the wind power in Qinghai Province during 1961—2022， “carbon peak” period of 2026—2035 and “carbon neutral” period of 2056—2065. The results show： （1） From 1961 to 2022， the average annual 10 m wind speed in Qinghai Province is 3.06 m·s-1， the 100 m wind power densities is 125.14 W·m-2， and the number of wind energy available days is 167. The wind power decreases gradually from the west to the east. The wind energy resources in Tsaidam Basin and Tanggula Mountains reach the suitable development standard， and the maximum value is in the west of Burhan Budai Mountains； （2） Wind power in Qinghai Province exhibits a decreasing trend. The reduction speed of 10 m wind speed is 2.08%·a-1， the 100 m wind power density is 1.40%·a-1 and the available days of wind energy is 0.88%·a-1； （3） Compared with the historical period， the 10 m wind speed at is 2.52 m·s-1 （2.39~2.63 m·s-1） with a decrease of 3.53% （2.34%~5.24%）， the wind power density is 50.28 W·m-2 （46.17~64.08 W·m-2） with a decrease of 10.20% （6.69%~14.91%）， and the wind energy available days is 114 days （105~124 days） with a decrease of 7.58% （2.44%~11.05%） from 2026 to 2035. From 2056 to 2065， the wind speed at 10 m is 2.49 m·s-1 （2.31~2.59 m·s-1） with a decrease of 5.71% （5.21%~6.53%）， the wind power density is 47.75 W·m-2 （38.89~54.79 W·m-2） with a decrease of 15.79% （14.72%~18.32%）， and the wind energy available days is 111 days （99~120 days） with a decrease of 10.73% （8.39%~13.55%）； （4） From 2026 to 2035 and 2056 to 2065， the 10 m wind speed at low emission scenarios （SSP1-1.9， SSP1-2.6 and SSP4-3.4） is 1.75% and 5.16% higher than high emission scenarios （SSP3-7.0， SSP5-8.5）， the 100 m wind power density is 5.23% and 14.28% higher， and the wind energy available days is 3.32% and 9.71% more. In general， the wind power in western part of Qinghai Province， especially Tsaidam Basin and Tanggula Mountains， is most abundant with relatively minor reduction， which has broad prospects for development and utilization.
Glaciers and snow are important components of the cryosphere. The subglacial topography， shallow structure of glaciers and snow， and snow thickness are of great significance for water resources investigation， glaciers and snow disaster early warning and global sea level prediction. Snow/ice detection radar has excellent penetration for semi-transparent media like glaciers or snow， and has become an important detection means in cryosphere fields. Radar tomography is a new technique for perspective three-dimensional imaging of snow and ice， with broad application prospects in key elements inversion such as snow/ice thickness and shallow structure. This paper first introduces the basic principles and methods of three kinds of tomography techniques based on ice radar and synthetic aperture radar （SAR）， and focuses on the research progress of the snow/ice penetrating radar tomography systems and their applications over the past 20 years on domestic and foreign ground-， air- and space-based platforms. The characteristics and applicability of three kinds of tomography techniques are summarized and analyzed， and the current challenges and development trends are discussed. This paper can provide references for the development of China’s snow/ice tomography radar system， the retrieval of global ice and snow key elements， and ice layer detection on the moon or Mars.
Glacier erosion and sedimentary landforms contain important ancient climate and environmental information， which are indicators for identifying the scope and nature of glacier action and provide a foundation for glacier dynamics research. After more than a hundred and fifty years of development， the expression research of glacier landforms has formed a relatively complete mapping scheme for glacier landforms， providing an important basis for determining the dynamic characteristics of the glacier formation process， glacier types， and glaciation. However， the professionalism of thematic maps of glacier landforms not only provides support for the reconstruction of ancient glaciers and the study of ancient environments， but also increases the difficulty of interpretation for non-experts. The extended map expression mechanism based on virtual reality/augmented reality maps， hybrid reality maps， micro maps， semantic maps， and metaphorical maps provides more dimensional expression solutions. On the basis of traditional spatial geometric dimensions， dimensions such as state and reader’s perspective are added， providing a new perspective for glacier landform expression. This article uses virtual reality （VR） technology to construct an interactive three-dimensional glacier landform scene of the Karola Glacier West Gully， adding dimensions such as three-dimensional， dynamic， and reader’s perspective. Based on previous research work （sediment characteristics， dating， etc.）， typical glacier landforms in the scene are annotated， and the research understanding of glacier landforms in the early stage is well matched with VR scenes. This scene can be freely transformed at 720 degrees， including 40 sub scenes from the Last Glacial Maximum to Late Glacier， Early Holocene， Neoglacial， Little Ice Age， and Modern Glacier， stimulating readers’ immersion， interaction， and conceptualization. This study shows that VR and other technologies can provide a extended map style representation of glacier landforms in more dimensions， which will further promote the reconstruction of ancient glaciers and the study of ancient climate and environment. It can provide good interactive three-dimensional glacier geomorphology materials for glacier geomorphology teaching， science popularization， research， and academic exchange， which can assist readers in more accurate identification of glacier relics and understanding of glacier development conditions.
