Permafrost is one of the crucial components of the cryosphere, covering about 25% of the global continental area. The active layer thickness (ALT), as the main site for heat and water exchange between permafrost and the external atmosphere, its changes significantly impact the carbon cycle, hydrological processes, ecosystems, and the safety of engineering structures in cold regions. This study constructs a Stefan CatBoost-ET (SCE) model through machine learning and Blending integration, leveraging multi-source remote sensing data, the Stefan equation, and measured ALT data to focus on the ALT in the Qinghai-Tibet Plateau (QTP). Additionally, the SCE model was verified via ten-fold cross-validation (MAE: 20.713 cm, RMSE: 32.680 cm, R2: 0.873, and MAPE: 0.104), and its inversion of QTP's ALT data from 1958 to 2022 revealed 1998 as a key turning point with a slow growth rate of 0.25 cm/a before 1998 and a significantly increased rate of 1.26 cm/a afterward. Finally, based on multiple model input factor analysis methods (SHAP, Pearson correlation, and Random Forest Importance), the study analyzed the ranking of key factors influencing ALT changes. Meanwhile, the importance of Stefan equation results in SCE model is verified. The research results of this paper have positive implications for eco-hydrology in the QTP region, and also provide valuable references for simulating the ALT of permafrost.

Climate change has been a strong driving force impacting the distribution of global water resources over the past few decades, especially in cold regions at high latitudes. Hydrological models are essential to analyse complex changing cold region's processes, such as permafrost, seasonally frozen soil, and snow cover, which are prevalent across much of Canada and the pan-Arctic basins. Here, we utilize the Hydrological Predictions for the Environment (HYPE) model with seven discretized vertical soil layers to assess climate change response to different water balance portioning components and permafrost extent. The study also explores seasonal and interannual shifts, examining the implications of model uncertainty associated with streamflow generation for the Nelson Churchill River Basin (NCRB). The calibrated HYPE model is run with a suite of fourteen GCMs and two RCPs (RCP 4.5 and RCP 8.5) scenarios representing 87% of the variability of 154 climate scenarios to discern the relationship between climate projections and water balance components. Increasing precipitation and temperature are anticipated in the future, but reduced, or balanced runoff is projected due to the dominant impact of rising temperature on evapotranspiration from thawing soil layers. Under an extreme scenario (RCP 8.5) 82% reduction in permafrost degradation is projected by the mid-future period (2050s). In this study, the future projections of streamflow, soil moisture, permafrost projection, and interrelationships of water balance processes at a continental scale are presented to aid in large-scale planning and implementation of sustainable development principles and guidelines for decision-making in the NCRB. Le changement climatique a & eacute;t & eacute; une force motrice majeure influen & ccedil;ant la r & eacute;partition des ressources en eau & agrave; l'& eacute;chelle mondiale au cours des derni & egrave;res d & eacute;cennies, en particulier dans les r & eacute;gions froides des hautes latitudes. Les mod & egrave;les hydrologiques sont essentiels pour analyser les processus complexes en & eacute;volution dans les r & eacute;gions froides, tels que le perg & eacute;lisol, les sols gel & eacute;s de mani & egrave;re saisonni & egrave;re et le couvert neigeux, qui sont r & eacute;pandus dans une grande partie du Canada et des bassins pan-arctiques. Dans cette & eacute;tude, nous utilisons le mod & egrave;le Hydrological Predictions for the Environment (HYPE), qui comprend sept couches de sol verticales discr & eacute;tis & eacute;es, pour & eacute;valuer la r & eacute;ponse au changement climatique des composantes du bilan hydrique et de l'& eacute;tendue du perg & eacute;lisol. L'& eacute;tude explore & eacute;galement les variations saisonni & egrave;res et interannuelles, en examinant les implications de l'incertitude du mod & egrave;le associ & eacute;e & agrave; la g & eacute;n & eacute;ration des d & eacute;bits fluviaux dans le bassin de la rivi & egrave;re Nelson Churchill (NCRB). Le mod & egrave;le HYPE calibr & eacute; est ex & eacute;cut & eacute; avec une s & eacute;rie de quatorze mod & egrave;les climatiques globaux (GCM) et deux sc & eacute;narios RCP (RCP 4.5 et RCP 8.5), repr & eacute;sentant 87 % de la variabilit & eacute; de 154 sc & eacute;narios climatiques, afin d'analyser la relation entre les projections climatiques et les composantes du bilan hydrique. Une augmentation des pr & eacute;cipitations et des temp & eacute;ratures est anticip & eacute;e dans le futur, mais un ruissellement r & eacute;duit ou & eacute;quilibr & eacute; est projet & eacute; en raison de l'impact dominant de la hausse des temp & eacute;ratures sur l'& eacute;vapotranspiration provenant des couches de sol en d & eacute;gel. Dans un sc & eacute;nario extr & ecirc;me (RCP 8.5), une r & eacute;duction de 82 % de la d & eacute;gradation du perg & eacute;lisol est projet & eacute;e d'ici la p & eacute;riode du milieu du si & egrave;cle (ann & eacute;es 2050). Cette & eacute;tude pr & eacute;sente des projections futures du d & eacute;bit fluvial, de l'humidit & eacute; du sol, de la d & eacute;gradation du perg & eacute;lisol et des interrelations des processus du bilan hydrique & agrave; l'& eacute;chelle continentale afin de soutenir la planification & agrave; grande & eacute;chelle et la mise en oeuvre de principes de d & eacute;veloppement durable pour & eacute;clairer la prise de d & eacute;cision dans le NCRB.

Permafrost roughly affects half of the boreal region in Alaska and varies greatly in its thermo-physical properties and genesis. In boreal ecosystems, permafrost formation and degradation respond to complex interactions among climate, topography, hydrology, soils, vegetation, and disturbance. We synthesized data on soil thermal conditions and permafrost characteristics to assess current permafrost conditions in central Alaska, and classified and mapped soil landscapes vulnerable to future thaw and thermokarst development. Permafrost soil properties at 160 sites ranged from rocky soils in hillslope colluvium and glacial till, to silty loess, to thick peats on abandoned floodplains and bogs, across 64 geomorphic units. Ground-ice contents (% moisture) varied greatly across geomorphic units. Mean annual ground temperatures at similar to 1 m depth varied 12.5 degrees C across 77 sites with most permafrost near thawing or actively thawing. To assess the vulnerability of permafrost to climate variability and disturbance, we differentiated permafrost responses in terms of rate of thaw, potential thaw settlement, and thermokarst development. Using a rule-based model that uses geomorphic units for spatial extrapolation at the landscape scale, we mapped 10 vulnerability classes across three areas in central Alaska ranging from high potential settlement/low thaw rate in extremely ice-rich loess to low potential settlement/high thaw rate in rocky hillslope colluvium. Permafrost degradation is expected to result in 10 thermokarst landform types. Vulnerability classes corresponded to thermokarst features that developed in response to past climates. Differing patterns in permafrost vulnerability have large implications for ecosystem trajectories, land use, and infrastructure damage from permafrost thaw.

The accelerated warming in the Arctic poses serious risks to freshwater ecosystems by altering streamflow and river thermal regimes. However, limited research on Arctic River water temperatures exists due to data scarcity and the absence of robust methodologies, which often focus on large, major river basins. To address this, we leveraged the newly released, extensive AKTEMP data set and advanced machine learning techniques to develop a Long Short-Term Memory (LSTM) model. By incorporating ERA5-Land reanalysis data and integrating physical understanding into data-driven processes, our model advanced river water temperature predictions in ungauged, snow- and permafrost-affected basins in Alaska. Our model outperformed existing approaches in high-latitude regions, achieving a median Nash-Sutcliffe Efficiency of 0.95 and root mean squared error of 1.0 degrees C. The LSTM model learned air temperature, soil temperature, solar radiation, and thermal radiation-factors associated with energy balance-were the most important drivers of river temperature dynamics. Soil moisture and snow water equivalent were highlighted as critical factors representing key processes such as thawing, melting, and groundwater contributions. Glaciers and permafrost were also identified as important covariates, particularly in seasonal river water temperature predictions. Our LSTM model successfully captured the complex relationships between hydrometeorological factors and river water temperatures across varying timescales and hydrological conditions. This scalable and transferable approach can be potentially applied across the Arctic, offering valuable insights for future conservation and management efforts.

Performance of pile foundations is significantly influenced by the micromechanical behaviour of the pile-soil interface. In frozen ground, the interface behaviour is temperature-dependent and viscous. This study examines the pullout behaviour of a single pile embedded in frozen soil using a two-dimensional axisymmetric finite element model developed with Abaqus software. A cohesive zone model (CZM) is employed along with a penalty type contact interaction to represent the interface behaviour while a time-hardening power-law creep model is used to simulate the viscous response of the interaction. The model is coupled with temperature and verified by data from temperature-controlled laboratory pullout and creep tests of model pile in frozen sand. The findings demonstrated that CZM effectively describes the behaviour of the frozen interface. Conceptual models that account for non-uniform temperature profiles and variable temperature time-series, showcase the significant influence of temperature on the overall performance of pile foundations in cold regions.

The degradation of subarctic peatland ecosystems under climate change impacts surrounding landscapes, carbon balance, and biogeochemical cycles. To assess these ecosystems' responses to climate change, it is essential to consider not only the active-layer thickness but also its thermo-hydraulic conditions. Ground-penetrating radar is one of the leading methods for studying the active layer, and this paper proposes systematically investigating its potential to determine the thermal properties of the active layer. Collected experimental data confirm temperature hysteresis in peat linked to changes in water and ice content, which GPR may detect. Using palsa mires of the Kola Peninsula (NW Russia) as a case study, we analyze relationships between peat parameters in the active layer and search for thermal gradient responses in GPR signal attributes. The results reveal that frequency-dependent GPR attributes can delineate thermal intervals of +/- 1 degrees C through disperse waveguides. However, further verification is needed to clarify the conditions under which GPR can reliably detect temperature variations in peat, considering factors such as moisture content and peat structure. In conclusion, our study discusses the potential of GPR for remotely monitoring freeze-thaw processes and moisture distribution in frozen peatlands and its role as a valuable tool for studying peat thermal properties in terms of permafrost stability prediction.

The fine-scale controls of active layer dynamics remain poorly understood, particularly at the southern boundary of continuous permafrost. We examined how environmental conditions associated with upland tundra heath, open graminoid fen, and palsa/peat plateau landforms affected active layer thermal regime (timing, magnitude, and rate of thaw) in a subarctic peatland in the Hudson Bay Lowlands, Canada. A significant increase in active layer thaw depth was evident between 2012 and 2024. Within-season thaw patterns differed among landforms, with tundra heath exhibiting the highest thaw rates and soil temperatures, succeeded by fen and palsa. Air temperature mediated by soil properties, topography, and vegetation affected thaw patterns. The increased thermal conductivity of gravel/sandy tundra heath soils exerted a more pronounced influence on thaw patterns relative to fens and palsas, both of which had a thicker organic layer. Near-surface soil moisture was the lowest in tundra, followed by palsas, and fens. Increased soil moisture impeded active layer thaw, likely due to a combination of soil surface evaporation and meltwater percolation. These findings elucidate the relationship between the biophysical properties of landform features and climate, revealing their role in influencing active layer thaw patterns in a subarctic ecosystem.

This study employs the Global Navigation Satellite System-Interferometric Reflectometry (GNSS-IR) technique, along with in situ hydrothermal data, to explore the details and mechanisms of permafrost ground surface deformation in the hinterland Tibetan Plateau. Through analyzing GNSS data collected from November 2021 to April 2024, seasonal deformation of up to approximately 5 cm, caused by active layer freeze-thaw cycles, was identified. Additionally, more than 2 years of continuous monitoring revealed a clear ground subsidence rate of 2.7 cm per year due to permafrost thawing. We compared the GNSS-IR monitored deformation with simulated deformation using in situ soil moisture and temperature profiles at 5-220 cm depth and found that the correlation reached 0.9 during the active-layer thawing and freezing period; we also observed that following an exceptionally thawing season, the subsequent thawing season experiences notably greater thaw subsidence. Furthermore, we analyzed the differences in GNSS-IR monitoring results with and without the inclusion of Beidou Navigation Satellite System (BDS) signals, and found that the inclusion of BDS signals reduced the standard deviation of GNSS-IR results by an average of 0.24 mm on snow-free periods, but increased by an average of 0.12 mm during the snow cover periods. This may be due to the longer wavelength of the BDS signal, which exhibits greater diffraction through snow and reduces signal reflectivity compared to other satellite systems. The research results demonstrate the potential and ability of continuous GNSS-IR ground surface deformation monitoring in revealing and exploring the hydrothermal processes within permafrost under climate change.

The extent of wildfires in tundra ecosystems has dramatically increased since the turn of the 21st century due to climate change and the resulting amplified Arctic warming. We simultaneously studied the recovery of vegetation, subsurface soil moisture, and active layer thickness (ALT) post-fire in the permafrost-underlain uplands of the Yukon-Kuskokwim Delta in southwestern Alaska to understand the interaction between these factors and their potential implications. We used a space-for-time substitution methodology with 2017 Landsat 8 imagery and synthetic aperture radar products, along with 2016 field data, to analyze tundra recovery trajectories in areas burned from 1953 to 2017. We found that spectral indices describing vegetation greenness and surface albedo in burned areas approached the unburned baseline within a decade post-fire, but ecological succession takes decades. ALT was higher in burned areas compared to unburned areas initially after the fire but negatively correlated with soil moisture. Soil moisture was significantly higher in burned areas than in unburned areas. Water table depth (WTD) was 10 cm shallower in burned areas, consistent with 10 cm of the surface organic layer burned off during fire. Soil moisture and WTD did not recover in the 46 years covered by this study and appear linked to the long recovery time of the organic layer.

The northwestern region of China is characterized by loess soil and seasonal permafrost. Due to the combined effects of its unique climate and precipitation patterns, local roads frequently suffer from issues such as foundation settlement, erosion, and collapse, which pose significant risks to both road construction and safe operation. This study examines a typical high subgrade in Northwest China, where a scaled laboratory model experiment was conducted. The research investigates the impact of water infiltration at the slope foot, under the dual influences of extreme cold and precipitation, on changes in the internal moisture field and settlement deformation characteristics of both the foundation and subgrade. The results indicate that the variation in moisture content across the follows an arc-shaped diffusion pattern. Settlement is influenced by both the amount of infiltrated water and cold air, with a noticeable lag effect. A settlement of 0.1 cm is considered the threshold for significant impact, with the minimum observed lag period approaching 4 days. The settlement is concentrated in the slope region, exhibiting a bending failure pattern. Numerical simulations reveal that the cross-sectional settlement distribution forms an inverted S shape, and the cumulative moisture content at each monitoring point exhibits a quadratic relationship with the cumulative settlement. The findings of this study provide scientific guidance and technical references for road construction and safe operation in the seasonal permafrost regions of Northwest China.