Svalbards permafrost is thawing as a direct consequence of climate change. In the Low Arctic, vegetation has been shown to slow down and reduce the active layer thaw, yet it is unknown whether this also applies to High Arctic regions like Svalbard where vegetation is smaller, sparser, and thus likely less able to insulate the soil. Therefore, it remains unknown which components of High Arctic vegetation impact active layer thaw and at which temporal scale this insulation could be effective. Such knowledge is necessary to predict and understand future changes in active layer in a changing Arctic. In this study we used frost tubes placed in study grids located in Svalbard with known vegetation composition, to monitor the progression of active layer thaw and analyze the relationship between vegetation composition, vegetation structure and snow conditions, and active layer thaw early in summer. We found that moss thickness, shrub and forb height, and vascular vegetation cover delayed soil thaw immediately after snow melt. These insulating effects attenuated as thaw progressed, until no effect on thaw depth was present after 8 weeks. High Arctic mosses are expected to decline due to climate change, which could lead to a loss in insulating capacity, potentially accelerating early summer active layer thaw. This may have important repercussions for a wide range of ecosystem functions such as plant phenology and decomposition processes. Temperatures are rising in the Arctic, causing increased thaw of the layer of soil located above the permanently frozen ground. In Low Arctic regions vegetation cools the soil, which reduces the thawing. So far, we do not know whether the small plants growing in the High Arctic may be able to slow or reduce thaw. We measured soil thaw throughout the summer in High Arctic Svalbard in locations where vegetation composition is known. We also measured thickness of the moss layer, height of plants and snow depth. We found that moss thickness was the strongest factor in insulating the soil. Also the cover of plants, height of shrubs and forbs, and height of grass-like plants slowed soil thaw in the early summer. The insulating effects became less over time and no effects were found 8 weeks after onset of thaw. As climate change is causing changes in the Arctic vegetation, mosses and small shrubs are expected to decrease. As we found these to be the most important factors in insulating the soil, a future decrease in mosses and small shrubs may cause accelerated soil thaw at the start of summer. High Arctic vegetation slows active layer thaw in early summer after snow melt Mosses show a stronger negative relation with thaw depth than vascular vegetation Factors influencing active layer thaw change over time in early summer

Improved modeling of permafrost active layer freeze-thaw plays a crucial role in understanding the response of the Arctic ecosystem to the accelerating warming trend in the region over the past decades. However, modeling the dynamics of the active layer at diurnal time scale remains challenging using the traditional models of freeze-thaw processes. In this study, a physically based analytical model is formulated to simulate the thaw depth of the active layer under changing boundary conditions of soil heat flux. Conservation of energy for the active layer leads to a nonlinear integral equation of the thaw depth using a temperature profile approximated from the analytical solution of the heat transfer equation forced by ground heat flux. Temporally variable ground heat flux is estimated using non-gradient models when field observations are not available. Validation of the proposed model conducted against field data obtained from three Arctic forest and tundra sites demonstrates that the model is able to simulate both thaw depth and soil temperature profiles accurately. The model has the potential to estimate regional variability of the thaw depth for permafrost related applications. The seasonally thawed layer on top of the permafrost (active layer) is a key component of the Arctic system affected by the strong warming trend over the past decades. This soil layer experiences a pronounced seasonal cycle of freezing and thawing processes caused by the availability of Sun's energy. Mathematical modeling of the thaw depth of the active layer has remained challenging. This study formulates a novel model for the simulation of the diurnal cycle of thawing process. The formulation is developed using innovative models of heat flux that goes into the soil and soil temperature profile. Ground heat flux is derived from available energy at the land surface using a theory of surface heat flux partition. The soil temperature profile is expressed using ground heat flux within the active layer. The proposed model has been validated against field observations during thawing season. The model simulation and field observations of the thaw depth are in a good agreement at three Arctic study sites with forest and tundra surface conditions. The proposed formulation can be used for modeling freeze-thaw cycles of the active layer at the regional scales since data on surface available energy can be obtained from remote sensing observations. The proposed model is highly effective in modeling thawing depth at higher time resolution and representing the soil energy budget Non-gradient models demonstrate a strong capability to model soil energy budget in data-sparse harsh environments

An accurate estimation of thaw depth is critical to understanding permafrost changes due to climate warming on the Qinghai-Tibetan Plateau (QTP). However, previous studies mainly focused on the interannual changes of active layer thickness (ALT) across the QTP, and little is known about the changes in the seasonal thaw depth. Machine learning (ML) is a critical tool to accurately estimate the ALT of permafrost, but a direct comparison of ML with deep learning (DL) in ALT projection regarding the model performance is still lacking. Here, ML, namely randomforest (RF), and DL algorithms like convolutional neural networks (CNN) and long short-term memory (LSTM) neural networks were compared to estimate the interannual changes of ALT and seasonal thaw depth on the QTP. Meteorological series, in-situ collected ALT observations, and geospatial information were used as predictors. The results show that both ML and DL methods are capable of estimating ALT and seasonal thaw depth in permafrost areas. The CNN and LSTM models developed using longer lagging times exhibit better performance in thaw depth prediction while the RF models are either mediocre or sometimes even worse as the lagging time increases. The results showthat the ALT from 2003 to 2011 on the QTP exhibits an increasing trend, especially in the northern region. In addition, 68.8%, 88.7%, 52.5%, and 47.5% of the permafrost regions on the QTP have deepened seasonal thaw depth in spring, summer, autumn, and winter, respectively. The correlation between air temperature and permafrost thaw depth ranges from 0.65 to 1 with the time lag ranging from 1 to 32 days. This study shows that ML and DL can be effectively used in retrieving ALT and seasonal thaw depth of permafrost, and could present an efficient way to figure out the interannual and seasonal variations of permafrost conditions under climate warming.

Climate change is destabilizing permafrost landscapes, affecting infrastructure, ecosystems, and human livelihoods. The rate of permafrost thaw is controlled by surface and subsurface properties and processes, all of which are potentially linked with each other. However, no standardized protocol exists for measuring permafrost thaw and related processes and properties in a linked manner. The permafrost thaw action group of the Terrestrial Multidisciplinary distributed Observatories for the Study of the Arctic Connections (T-MOSAiC) project has developed a protocol, for use by non-specialist scientists and technicians, citizen scientists, and indigenous groups, to collect standardized metadata and data on permafrost thaw. The protocol introduced here addresses the need to jointly measure permafrost thaw and the associated surface and subsurface environmental conditions. The parameters measured along transects include: snow depth, thaw depth, vegetation height, soil texture, and water level. The metadata collection includes data on timing of data collection, geographical coordinates, land surface characteristics (vegetation, ground surface, water conditions), as well as photographs. Our hope is that this openly available dataset will also be highly valuable for validation and parameterization of numerical and conceptual models, and thus to the broad community represented by the T-MOSAiC project.

The spatial distribution of permafrost and associated mean annual ground temperature (MAGT) and active layer thickness (ALT) are crucial data for hydrological studies. In this paper, we present the current state of knowledge on the spatial distribution of the permafrost properties of 29 river basins in Mongolia. The MAGT and ALT values are estimated by applying TTOP and Kudryavtsev methods. The main input of both methods is the spatially distributed surface temperature. We used the 8-day land surface temperature (LST) data from the day- and night-time Aqua and Terra images of the moderate resolution imaging spectroradiometer (MODIS). The gaps of the MODIS LST data were filled by spatial interpolation. Next, an LST model was developed based on 34 observational borehole data using a panel regression analysis (Baltagi, Econometric analysis of panel data, 3 edn, Wiley, New York, 2005). The model was applied for the whole country and covered the period from August 2012 to August 2013. The results show that the permafrost covers 26.3% of the country. The average MAGT and ALT for the permafrost region is - 1.6 degrees C and 3.1 m, respectively. The MAGT above -2 degrees C (warm permafrost) covers approximately 67% of the total permafrost area. The permafrost area and distribution in cold and warm permafrost varies highly over the country, in particular in regions where the river network is highly developed. High surface temperatures associated with climate change would result in changes of permafrost conditions, and, thus, would impact the surface water availability in these regions. The data on permafrost conditions presented in this paper can be used for further research on changes in the hydrological conditions of Mongolia.

This paper presents the results of 39 years of observations conducted at the Chabyda station to monitor the thermal state of permafrost landscapes under current climatic warming. The analysis of long-term records from weather stations in the region has revealed one of the highest increasing trends in mean annual air temperature in northern Russia. The partitioning of the energy balance in different landscape units within the study area has been analyzed. Quantitative relationships in the long-term variability of ground thermal parameters, such as the ground temperature at the bottom of the active layer and seasonal thaw depth, have been established. The ground temperature dynamics within the depth of zero annual amplitude indicates that both warm and cold permafrost are thermally stable. The short-term variability of the snow accumulation regime is the main factor controlling the thermal state of the ground in permafrost landscapes. The depth of seasonal thaw is characterized by low interannual variability and exhibits little response to climate warming, with no statistically significant increasing or decreasing trend. The results of the ground thermal monitoring can be extended to similar landscapes in the region, providing a reliable basis for predicting heat transfer in natural, undisturbed landscapes.

Aims For informed predictions on the sensitivity of Arctic tundra landscape to permafrost thaw, we aimed to investigate the distribution pattern of near-surface ground ice and its influencing factors in Northeast Siberia. Methods Near-surface permafrost cores (60 cm) were sampled along small-scale topographic gradients in two drained lakebeds. We investigated which factors (vegetation, hydrological and soil) correlated strongest with ice content and explored its spatial heterogeneity at different scales (1 to 100 m). Results The ice content was highest in the depressions of the wet lakebed and lowest at the slopes of the dry lakebed. In the wet lakebed the ice content increased with depth, while in the dry lakebed the vertical distribution depended on topographical position. Spatial variability in ice content was similar at different scales, stressing strong influence of local drivers. 0-60 cm ice content correlated strongest with soil moisture of the overlying unfrozen soil, while 0-20 cm ice content correlated strongest with vegetation characteristics. Conclusions Our study implies that vegetation effect on microclimate is strong enough to affect near-surface ice distribution, and that ice-rich tundra may be highly sensitive to thaw once climate warming offsets the protective impact of vegetation.

Terrestrial Arctic ecosystems play a key role in the global carbon (C) cycle, as they store a large amount of organic matter in permafrost. Among regions with continuous permafrost, Svalbard has one of the warmest permafrost and may provide a template of the environmental responses of Arctic regions to future climate change. We analyze the CO2 fluxes at a polygonal tundra site in Adventdalen (Svalbard) during one full growing season across a vegetation and environmental gradient to understand how the interaction of different abiotic (thaw depth, ground surface temperature (GST), soil moisture, photosynthetic active radiation - PAR) and biotic (leaf area index (LAI), and plant phenology) factors affect the CO2 fluxes and identify the drivers of Net Ecosystem Exchange (NEE) and Ecosystem Respiration (ER). Three distinct periods (early, peak, and late) characterized the growing season based on plant phenology and the main environmental conditions. Comparing early, peak and late season, both NEE and ER exhibited specific patterns: ER shown high values since the early season, only slightly increased at peak, and then decreased drastically in the late season, with GST being the most important driver of ER. The drivers of NEE changed during the season: thaw depth, PAR and GST during the early season, LAI at peak, and PAR during the late season. These data allow to highlight that the thawing and freezing of the upper part of the active layer during the early and late season controls ER, possibly due to the response of microbial respiration in the upper part of the soil. Especially during the late season, despite the fully developed active layer (reaching its highest thawing depth), the freezing of the uppermost 2 cm of soil induced the drastic decrease of the respiratory efflux. In addition, the seasonal C balance of our plots indicated a seasonal source at our plots, in apparent contrast with previous eddy covariance (EC) measurements from a wetter area nearby. This difference implies that drier ecosystems act as sources while wetter ecosystems are sinks, suggesting that a drying trend in polygonal tundra could switch these ecosystems from CO2 sinks to sources in a feedback to future climate change.

A new Circumpolar Active Layer Monitoring (CALM) site was established in 2009 at the Limnopolar Lake watershed in Byers Peninsula, Livingston Island, Antarctica, to provide a node in the western Antarctic Peninsula, one of the regions that recorded the highest air temperature increase in the planet during the last decades. The first detailed analysis of the temporal and spatial evolution of the thaw depth at the Limnopolar Lake CALM-S site is presented here, after eight years of monitoring. The average values range between 48 and 29 cm, decreasing at a ratio of 16 cm/decade. The annual thaw depth observations in the 100 x 100 m CALM grid are variable (Variability Index of 34 to 51%), although both the Variance Coefficient and the Climate Matrix Analysis Residual point to the internal consistency of the data. Those differences could be explained then by the terrain complexity and node-specific variability due to the ground properties. The interannual variability was about 60% during 2009-2012, increasing to 124% due to the presence of snow in 2013, 2015 and 2016. The snow has been proposed here as one of the most important factors controlling the spatial variability of ground thaw depth, since its values correlate with the snow thickness but also with the ground surface temperature and unconfined compression resistance, as measured in 2010. The topography explains the thaw depth spatial distribution pattern, being related to snowmelt water and its accumulation in low-elevation areas (downslope-flow). Patterned grounds and other surface features correlate well with high thaw depth patterns as well. The edaphic factor (E = 0.05842 m(2)/degrees C.day; R-2 = 0.63) is in agreement with other permafrost environments, since frozen index (F > N 0.67) and MAAT (<-2 degrees C) denote a continuous permafrost existence in the area. All these characteristics provided the basis for further comparative analyses between others nearby CALM sites. (C) 2017 Elsevier B.V. All rights reserved.

The response of dissolved organic carbon (DOC) flux to permafrost degradation is one of the major sources of uncertainty in predicting the permafrost carbon feedback. We investigated DOC export and properties over two complete flow seasons in a catchment on the northern Qinghai-Tibetan Plateau. DOC concentration and biodegradability decreased systematically as thaw depth increased through the season, attributable to changing carbon sources and degree of microbial processing. Increasing DOC aromaticity and C-13-DOC indicated shifts toward more recalcitrant carbon sources and greater residence time in soils prior to reaching the stream network. These strong and consistent seasonal trends suggest that gradual active layer deepening may decrease DOC export and biodegradability from permafrost catchments. Because these patterns are opposite observations from areas experiencing abrupt permafrost collapse (thermokarst), the overall impact of permafrost degradation on DOC flux and biodegradability may depend on the proportion of the landscape experiencing gradual thaw versus thermokarst.