Search for: All records

Abstract. Climate warming in the Arctic results in thawing permafrost and associated processes like thermokarst, especially in ice-rich permafrost regions. Since permafrost soils are one of the largest organic carbon reservoirs of the world, their thawing leads to the release of greenhouse gases due to increasing microbial activity with rising soil temperature, further exacerbating climate warming. To enhance the predictions of potential future impacts of permafrost thaw, a detailed assessment of changes in soil characteristics in response to thermokarst processes in permafrost landscapes is needed, which we investigated in this study in an Arctic coastal lowland. We analysed six sediment cores from the Arctic Coastal Plain of northern Alaska, each representing a different landscape feature along a gradient from upland to thermokarst lake and drained basin to thermokarst lagoon in various development stages. For the analysis, a multiproxy approach was used, including sedimentological (grain size, bulk density, ice content), biogeochemical (total organic carbon (TOC), TOC density (TOCvol), total nitrogen (TN), stable carbon isotopes (δ13C), TOC/TN ratio, mercury (Hg)), and lipid biomarker (n-alkanes, n-alkanols, and their ratios) parameters. We found that a semi-drained state of thermokarst lakes features the lowest OC content, and TOC and TN are generally higher in unfrozen deposits, hinting at a more intact state of organic matter. Indicated by the average chain length (ACL), δ13C, Paq, and Pwax, we found a stronger influence of aquatic organic matter (OM) in the OM composition in the soils covered by water compared to those not covered by water. Moreover, the results of the δ13C, TOC/TN ratio, and CPI indicate that the saline deposits contain stronger degraded OM than the deposits not influenced by saltwater. Additionally, we found positive correlations between the TOC and TOCvol and the Hg content in the deposits. The results indicate that thermokarst-influenced deposits tend to accumulate Hg during thawed periods and thus contain more Hg than the upland permafrost deposits that have not been impacted by lake formation. Our findings offer valuable insights into the dynamics of carbon storage and vulnerability to decomposition in coastal permafrost landscapes, reflecting the interplay of environmental factors, landform characteristics, and climate change impacts on Arctic permafrost environments. 

This dataset, called DARTS, comprises footprints of retrogressive thaw slump (RTS) identified and quantified using an automated deep learning approach in RTS hotspots across the Arctic and Subarctic permafrost regions. We utilized multispectral PlanetScope imagery with a spatial resolution of ~3 meters (m), complemented by ArcticDEM (Digital Elevation Model) and derived datasets, including slope, relative elevation, and Landsat-derived change trends. The dataset covers an area of 1.6 million square-kilometers (km²), with at least one coverage between 2021 and 2023, and provides annual coverage for approximately 900,000 km². In several highly active key sites, such as Banks Island, Peel Plateau, and Novaya Zemlya, we extended the data frequency and temporal coverage to 2018-2023. We mapped a total of more than 43,000 individual RTS and ALD, many of them multiple times. We offer two levels of datasets; Level 1: RTS footprints per image with timestamps; and Level 2: annually aggregated RTS footprints. Essential metadata includes image footprints, dataset coverage, timestamps, and model-specific information. To enhance reproducibility and further use, the training labels, processing code, and model checkpoints are publicly available. This version, v1.1, is the revised first openly accessible release. The dataset will be maintained and continuously updated in both spatial and temporal extent. It can be used for mapping and quantifying RTS, analyzing spatio-temporal patterns of RTS dynamics, or serving as input for landscape dynamics models. 

The rapid climate warming is affecting the Arctic which is rich in aquatic systems. As a result of permafrost thaw, thermokarst lakes and ponds are either shrinking due to lake drainage or expanding due to lake shore erosion. This process in turn mobilizes organic carbon, which is released by permafrost deposits and active layer material that slips into the lake. In this study, we combine hydrochemical measurements and remote sensing data to analyze the influence of lake change processes, especially lake growth, on lake hydrochemical parameters such as DOC, EC, pH as well as stable oxygen and hydrogen isotopes in the Arctic Coastal Plain. For our entire dataset of 97 water samples from 82 water bodies, we found significantly higher CH4 concentrations in lakes with a floating-ice regime and significantly higher DOC concentrations in lakes with a bedfast-ice regime. We show significantly lower CH4 concentrations in lagoons compared to lakes as a result of an effective CH4 oxidation that increased with a seawater connection. For our detailed lake sampling of two thermokarst lakes, we found a significant positive correlation for lake shore erosion and DOC for one of the lakes. Our detailed lake sampling approach indicates that the generally shallow thermokarst lakes are overall well mixed and that single hydrochemical samples are representative for the entire lake. Finally, our study confirms that DOC concentrations correlates with lake size, ecoregion type and underlying deposits. 

Arctic permafrost is facing significant changes due to global climate change. As these regions are largely inaccessible, remote sensing plays a crucial rule in better understanding the underlying processes across the Arctic. In this study, we focus on the remote detection of retrogressive thaw slumps (RTSs), a permafrost disturbance comparable to slow landslides. For such remote sensing tasks, deep learning has become an indispensable tool, but limited labeled training data remains a challenge for training accurate models. We present PixelDINO, a semi-supervised learning approach, to improve model generalization across the Arctic with a limited number of labels. PixelDINO leverages unlabeled data by training the model to define its own segmentation categories (pseudoclasses), promoting consistent structural learning across strong data augmentations. This allows the model to extract structural information from unlabeled data, supplementing the learning from labeled data. PixelDINO surpasses both supervised baselines and existing semi-supervised methods, achieving average intersection-over-union (IoU) of 30.2 and 39.5 on the two evaluation sets, representing significant improvements of 13% and 21%, respectively, over the strongest existing models. This highlights the potential for training robust models that generalize well to regions that were not included in the training data. 

Large proglacial lakes could have been a significant methane source during the last deglaciation. Today, proglacial lakes are small and mostly limited in the northern hemisphere to the margins of ice sheets in Greenland, Alaska, and Canada, but much larger proglacial lakes collectively flooded millions of square kilometers in the northern hemisphere over the last deglacial period. We synthesize new and existing methane flux measurements from modern proglacial lakes in Alaska and Greenland and use these data together with reconstructed lake area and bathymetry, new paleorecords of sediment organic geochemistry, carbon accumulation, and other proxies to broadly constrain the possible deglacial methane dynamics of a single large North American proglacial lake, Lake Agassiz. While large influxes of glaciogenic material contributed to rapid organic carbon burial during initial lakes phases, limited bioavailability of this carbon is suggested by its likely subglacial origin and prior microbial processing. Water depths of >20 m across 37–90% of the lake area facilitating significant oxidation of methane within the water column further limited emissions. Later phases of lake lowering and subsequent re-expansion into shallow aquatic and subaerial environments provided the most significant opportunity for methane production according to our estimates. We found that Lake Agassiz was likely a small source [0.4–2.7 Tg yr−1 mean (0.1–9.9 Tg yr−1 95% CI)] of methane during the last deglaciation on par with emissions from modern wildfires. Although poor constraints of past global proglacial lake areas and morphologies currently prevent extrapolation of our results, we suggest that these systems were likely an additional source of methane during the last deglacial transition that require further study. 

Atmospheric methane (CH4) concentrations have gone through rapid changes since the last deglaciation; however, the reasons for abrupt increases around 14,700 and 11,600 years before present (yrs BP) are not fully understood. Concurrent with deglaciation, sea-level rise gradually inundated vast areas of the low-lying Beringian shelf. This transformation of what was once a terrestrial-permafrost tundra-steppe landscape, into coastal, and subsequently, marine environments led to new sources of CH4 from the region to the atmosphere. Here, we estimate, based on an extended geospatial analysis, the area of Beringian coastal wetlands in 1000-year intervals and their potential contribution to northern CH4 flux (based on present day CH4 fluxes from coastal wetland) during the past 20,000 years. At its maximum (∼14,000 yrs BP) we estimated CH4 fluxes from Beringia coastal wetlands to be 3.5 (+4.0/-1.9) Tg CH4 yr−1. This shifts the onset of CH4 fluxes from northern regions earlier, towards the Bølling-Allerød, preceding peak emissions from the formation of northern high latitude thermokarst lakes and wetlands. Emissions associated with the inundation of Beringian coastal wetlands better align with polar ice core reconstructions of northern hemisphere sources of atmospheric CH4 during the last deglaciation, suggesting a connection between rising sea level, coastal wetland expansion, and enhanced CH4 emissions. 

Abstract High-latitude and altitude cold regions are affected by climate warming and permafrost degradation. One of the major concerns associated with degrading permafrost is thaw subsidence (TS) due to melting of excess ground ice and associated thaw consolidation. Field observations, remote sensing, and numerical modeling are used to measure and estimate the extent and rates of TS across broad spatial and temporal scales. Our new data synthesis effort from diverse permafrost regions of North America and Eurasia, confirms widespread TS across the panarctic permafrost domain with rates of up to 2 cm yr⁻¹in the areas with low ice content and more than 3 cm yr⁻¹in regions with ice-rich permafrost. Areas with human activities or areas affected by wildfires exhibited higher subsidence rates. Our findings suggest that permafrost landscapes are undergoing geomorphic change that is impacting hydrology, ecosystems, and human infrastructure. The development of a systematic TS monitoring is urgently needed to deliver consistent and continuous exchange of data across different permafrost regions. Integration of coordinated field observations, remote sensing, and modeling of TS across a range of scales would contribute to better understanding of rapidly changing permafrost environments and resulting climate feedbacks. 

This dataset, called DARTS, comprises footprints of retrogressive thaw slump (RTS) and active layer detachments slides (ALD) identified and quantified using an automated deep learning approach in RTS hotspots across the Arctic and Subarctic permafrost regions. We utilized multispectral PlanetScope imagery with a spatial resolution of 3 meters (m), complemented by ArcticDEM (Digital Elevation Models) and derived datasets, including slope, relative elevation, and Landsat-derived change trends. The dataset covers an area of 1.6 million square-kilometers (km²), with at least one coverage between 2021 and 2023, and provides annual coverage for approximately 900,000 km². In several highly active key sites, such as Banks Island, Peel Plateau, and Novaya Zemlya, we extended the data frequency and temporal coverage to 2018-2023. We mapped a total of more than 43,000 individual RTS and ALD, many of them multiple times. We offer two levels of datasets; Level 1: RTS footprints per image with timestamps; and Level 2: annually aggregated RTS footprints. Essential metadata includes image footprints, dataset coverage, timestamps, and model-specific information. To enhance reproducibility and further use, the training labels, processing code, and model checkpoints are publicly available. This version, v1, is the first openly accessible release. The dataset will be maintained and continuously updated in both spatial and temporal extent. It can be used for mapping and quantifying RTS, analyzing spatio-temporal patterns of RTS dynamics, or serving as input for landscape dynamics models.