Search for: All records

This dataset, called DARTS, comprises footprints of active parts 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 1.9 million square-kilometers (km²), with at least one coverage between 2016 and 2025, and provides annual coverage for approximately 570,000 km² between 2021 and 2025. In several highly active key sites, such as Banks Island, Peel Plateau, and Novaya Zemlya, we extended the data frequency and temporal coverage to before 2021. We mapped a total of more than 58,000 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, v1.3, is the updated first openly accessible release, extending version v1.2, which covered years 2018 through 2023. 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. 

### Access Data are available for download at [http://arcticdata.io/data/10.18739/A28G8FK10](http://arcticdata.io/data/10.18739/A28G8FK10) To download all files in the command line, run the following command in a terminal: `wget -r -np -nH --cut-dirs=3 -R '\?C=' -R robots.txt https://arcticdata.io/data/10.18739/A28G8FK10` To download only a subdirectory of the archived files, add the subdirectory to the end of the URL above. ### Overview Lake area for permanent and seasonal in the Arctic were derived from Landsat satellite imagery from 1984-2021 using machine learning. Data was collected for lakes with a maximum extent larger than one hectare, buffered by 30 meters, within four discontiguous transects across the Arctic region at latitudes greater than 60 degrees North. This region covers large parts of the northern permafrost terrain. This annual lake area data was merged with lake size trend attributes derived from the years 2011-2020. Lake area change rates were normalized by area and/or perimeter. The lake extraction methodology is based on Nitze et al, 2017 and 2018. This dataset is version 2.0. You can find version 1.0, related to Nitze et al. 2018 here: https://apgc.awi.de/dataset/hot-t1-prd-lake-ls-1999-2014 https://apgc.awi.de/dataset/hot-t2-prd-lake-ls-1999-2014 https://apgc.awi.de/dataset/hot-t3-prd-lake-ls-1999-2014 https://apgc.awi.de/dataset/hot-t4-prd-lake-ls-1999-2014 Data files were merged, and the annual permanent and seasonal water data for 2017-2021, a subset of the available years, was fed into the Permafrost Discovery Gateway visualization workflow. This workflow cleaned, standardized, and visualized the data and output two Tile Matrix Sets per year. One Tile Matrix Set is the data in the form of GeoPackages, or staged tiles, and the other Tile Matrix Set is the staged tiles in the form of GeoTIFF tiles. The highest resolution tiles were resampled to produce GeoTIFFs for lower resolutions. In the future, this dataset will be expanded for annual data for 1984-2021. This visualized data will be published on the Permafrost Discovery Gateway portal: https://arcticdata.io/catalog/portals/permafrost/Imagery-Viewer Data limitations in use: This data is part of an initial release of the pan-Arctic data product for lake area and size trends, and it is expected that there are constraints on its accuracy and completeness. Users are encouraged to provide feedback regarding how they use this data and issues they encounter during post-processing. Please reach out to the dataset contact or a member of the Permafrost Discovery Gateway team via support@arcticdata.io. 

Abstract Retrogressive Thaw Slumps (RTS) are widespread mass-wasting hillslope failures triggered by thawing permafrost. While regional studies have provided insights into the spatial distribution and dynamics of RTS, a consistent and unbiased quantification and monitoring remains unsolved at pan-arctic scales. We present the Database of AI-detected Arctic RTS footprints (DARTS), comprising ~43,000 individual footprints of active RTS or active areas within larger RTS landforms. DARTS spans ~1.6 million km²from 2018–2023, with at least annual coverage from 2021–2023 across a ~900,000 km²region. The database is freely available in two processing levels: sub-annual and annually aggregated polygon footprints including spatial and tabular metadata. DARTS uses a highly automated workflow based on deep learning segmentation of PlanetScope multi-spectral satellite imagery (3–5 m resolution) and elevation data. Validation against different regional RTS datasets yielded F1 scores ranging from 0.263 to 0.700, with higher accuracy in areas of intense RTS activity. DARTS provides a valuable resource for systematically mapping, quantifying, and analyzing active hillslope thermokarst distribution and changes over time across the circum-arctic permafrost region. 

Abstract Arctic permafrost is undergoing rapid changes due to climate warming in high latitudes. Retrogressive thaw slumps (RTS) are one of the most abrupt and impactful thermal-denudation events that change Arctic landscapes and accelerate carbon feedbacks. Their spatial distribution remains poorly characterised due to time-intensive conventional mapping methods. While numerous RTS studies have published standalone digitisation datasets, the lack of a centralised, unified database has limited their utilisation, affecting the scale of RTS studies and the generalisation ability of deep learning models. To address this, we established the Arctic Retrogressive Thaw Slumps (ARTS) dataset containing 23,529 RTS-present and 20,434 RTS-absent digitisations from 20 standalone datasets. We also proposed a Data Curation Framework as a working standard for RTS digitisations. This dataset is designed to be comprehensive, accessible, contributable, and adaptable for various RTS-related studies. This dataset and its accompanying curation framework establish a foundation for enhanced collaboration in RTS research, facilitating standardised data sharing and comprehensive analyses across the Arctic permafrost research community. 

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. 

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. 

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. 

Abstract. Permafrost thaw disrupts ecosystems, hydrology, and biogeochemical cycles, reinforcing climate change through a positive permafrost-carbon feedback loop. Thaw can be gradual, deepening the active layer, or abrupt, triggering thermokarst, thermo-erosion, or thermodenudation. Retrogressive thaw slumps (RTSs) are a key manifestation of abrupt permafrost thaw. Yet, their distribution, scale, and environmental controls in the West Siberian Arctic remain poorly understood, further complicated by their rapid evolution. This study presents an extensive update of the West Siberian RTS inventory through manual mapping using high-resolution, multi-source, multi-year recent (2016–2023) satellite basemaps (ESRI, Google Earth, and Yandex Maps). We developed an RTS classification capturing key environmental parameters, including morphology, spatial organization, terrain position, and associated relief-forming concurrent processes. The dataset comprises 6168 classified RTS landforms, integrating newly mapped sites with previously reported occurrences to provide a comprehensive view of a 445 226 km2 region covering the Yamal, Gydan, and Tazovsky peninsulas. The collected data underwent manual filtering and verification, leveraging local field experience and observations from key sites to reduce uncertainty and minimize false positives. Accuracy analysis, performed by comparing the dataset with various field datasets collected across the peninsulas, confirmed high accuracy (>90 %) for RTS identification. The dataset likely underestimated the distribution of small RTSs due to the resolution limitations of remote sensing data, hence generally providing a conservative estimate. This dataset serves as a valuable resource for diverse research fields, including ecology, biogeochemistry, geomorphology, climatology, permafrost science, and natural hazard assessment. Additionally, it provides a crucial reference dataset for machine learning applications, enhancing upcoming remote sensing classification and predictive modeling approaches. The dataset is available from Nesterova et al. (2025; https://doi.org/10.1594/PANGAEA.974406).