Search Results (24)

Retrogressive thaw slumps (RTSs) serve as key indicators of global climate change and also pose significant risks to critical infrastructure along the Qinghai–Tibet Engineering Corridor (QTEC). Accurate automatic segmentation of RTSs using Sentinel-2 imagery is of great value for climate change research and risk assessment, owing to the dataset’s ready availability and extensive spatiotemporal coverage. However, this segmentation task remains challenging due to the complex morphology and variable sizes of RTSs, as well as their low contrast and fuzzy boundaries against the surrounding landscape in medium-resolution satellite imagery. To deal with these challenges, this study proposes the Multi-Scale Object-aware Context Attention Network (MOCA-Net), which enhances the Swin Transformer backbone through two critical components: the Feature Enhancement Network and Enhanced Decoder. Evaluation metrics show that MOCA-Net outperforms seven mainstream baseline models, achieving a Mean Intersection over Union (mIoU) of 0.8609 and an RTS-class IoU of 0.7473. The qualitative visual evaluation further confirms MOCA-Net’s improved performance in delineating RTSs through more accurate morphologies and boundaries. Ablation studies confirm that each designed component contributes to the MOCA-Net’s segmentation performance, and their combination yields more balanced results. This model unlocks the capability of Sentinel-2 imagery for accurate RTS segmentation, making it promising for applications over large spatiotemporal extents. Full article

Despite the ecological sensitivity of the Yangtze River Source Region (YRSR), the current research critically lacks a quantified assessment of the spatial occurrence probability of Retrogressive Thaw Slumps (RTSs) in this specific high-altitude terrain. This study aims to bridge this knowledge gap by establishing a robust susceptibility assessment framework to accurately model the spatial distribution and risk levels of RTSs. The innovations of this research include (i) the construction of a complete and up-to-date 2024 RTS inventory for the entire YRSR based on high-resolution optical remote sensing; (ii) the integration of time-series spectral features (e.g., vegetation and moisture trends) alongside static topographic variables to enhance the physical interpretability of machine learning models; and (iii) the application of advanced ensemble learning algorithms combined with SHAP analysis to establish a comprehensive RTS susceptibility zonation. The results reveal a rapid intensification of instability, evidenced by an 83.5% surge in RTS abundance, with the CatBoost model achieving exceptional accuracy (AUC = 0.994), and identifying that specific static topographic factors (particularly elevations between 4693 and 4812 m and north-to-east aspect) and dynamic spectral anomalies (indicated by declining vegetation vigor and increasing surface wetness) are the dominant drivers controlling RTS distribution. This study provides essential baseline data and spatial guidance for ecological conservation and engineering maintenance in the Asian Water Tower, demonstrating a highly effective paradigm for monitoring permafrost hazards under climate warming. Full article

Climate warming is accelerating the degradation of permafrost, particularly in mid- to low-latitude regions, resulting in the widespread formation of thermokarst landscapes, including retrogressive thaw slumps (RTSs). These landforms, which are predominantly formed by the thawing of ice-rich permafrost, have been shown to impact topography, hydrology, and ecosystem dynamics. However, spatiotemporal changes in RTS distribution and development in mid- to low-latitude permafrost regions are not well understood. This study investigates RTS spatiotemporal dynamics in the Heshenling area of the western Qilian Mountains using multi-temporal PlanetScope and Google Earth imagery, along with Sentinel-1 InSAR data acquired from 2014 to 2023. The results reveal 20 RTSs, averaging 3.7 ha in area, primarily distributed on slopes of 7–23° and at elevations of 3455–3651 m a.s.l. The deformation rates of RTSs ranged from −54 to 27 mm/year. Three developmental stages—active, stable, and mature—were identified through analysis of surface deformation and geometric variations. Active RTSs exhibited accelerated headscarp retreat and debris tongue expansion, with some slumps expanding by up to 35%. This study highlights high temperatures and rainfall as potential factors contributing to the accelerated development of RTS in arid alpine environments, and suggests that RTS activity is likely to accelerate with continued climate change. Full article

Climate warming can lead to permafrost degradation, potentially resulting in slope failures such as retrogressive thaw slumps (RTSs). The formation of and changes in RTSs could exacerbate the degradation of permafrost and the environment in general. The mechanisms of RTS progression and the potential consequences on the analogous freeze–thaw cycle are not well understood, owing partly to necessitating field work under harsh conditions and with high costs. Here, we used multi-source remote sensing and field surveys to quantify the changes in an RTS on Eboling Mountain in the Qilian Mountain Range in west-central China. Based on optical remote sensing and SBAS-InSAR measurements, we analyzed the RTS evolution and the underlying drivers, combined with meteorological observations. The RTS expanded from 56 m² in 2015 to 4294 m² in 2022, growing at a rate of 1300 m²/a to its maximum in 2018 and then decreasing. Changes in temperature and precipitation play a dominant role in the evolution of the RTS, and the extreme weather in 2016 may also be a primary contributor to the accelerated growth, with an average deformation of −8.3 mm during the thawing period, which decreased slope stability. The RTS evolved more actively during the thawing and early freezing process, with earthquakes having potentially contributed further to RTS evolution. We anticipate that the rate of RTS evolution is likely to increase in the coming years. Full article

Retrogressive thaw slumps (RTS) are a form of abrupt permafrost thaw that can rapidly mobilize ancient frozen soil carbon, magnifying the permafrost carbon feedback. However, the magnitude of this effect is uncertain, largely due to limited information about the distribution and extent of RTS across the circumpolar region. Although deep learning methods such as Convolutional Neural Networks (CNN) have shown the ability to map RTS from high-resolution satellite imagery (≤10 m), challenges remain in deploying these models across large areas. Imagery selection and procurement remain one of the largest challenges to upscaling RTS mapping projects, as the user must balance cost with resolution and sensor quality. In this study, we compared the performance of three satellite imagery sources that differed in terms of sensor quality and cost in predicting RTS using a Unet3+ CNN model and identified RTS characteristics that impact detectability. Maxar WorldView imagery was the most expensive option, with a ground sample distance of 1.85 m in the multispectral bands (downloaded at 4 m resolution). Planet Labs PlanetScope imagery was a less expensive option with a ground sample distance of approximately 3.0–4.2 m (downloaded at 3 m resolution). Although PlanetScope imagery was downloaded at a higher resolution than WorldView, the radiometric footprint is around 10–12 m, resulting in less crisp imagery. Finally, Sentinel-2 imagery is freely available and has a 10 m resolution. We used 756 RTS polygons from seven sites across Arctic Canada and Siberia in model training and 63 RTS polygons in model testing. The mean IoU of the validation and testing data sets were 0.69 and 0.75 for the WorldView model, 0.70 and 0.71 for the PlanetScope model, and 0.66 and 0.68 for the Sentinel-2 model, respectively. The IoU of the RTS class was nonlinearly related to the RTS Area, showing a strong positive correlation that attenuated as the RTS Area increased. The models were better able to predict RTS that appeared bright on a dark background and were less able to predict RTS that had higher plant cover, indicating that bare ground was a primary way the models detected RTS. Additionally, the models performed less well in wet areas or areas with patchy ground cover. These results indicate that all imagery sources tested here were able to predict larger RTS, but higher-quality imagery allows more accurate detection of smaller RTS. Full article

The increasing incidence of retrogressive thaw slumps (RTSs) in permafrost regions underscores the need for detailed spatial and temporal analysis using InSAR technology to monitor and predict dynamic changes in the process of RTSs. Nevertheless, current InSAR deformation forecasting methods employing deep learning strategies such as the traditional long short-term memory (LSTM) and recent transformer models encounter difficulties in effectively capturing temporal features. Moreover, they are limited in their ability to directly integrate spatial information. In this paper, an innovative deep learning approach named Spacetimeformer is proposed for predicting medium- and short-term InSAR deformation of RTSs in the Chumar River area. This method employs a transformer architecture with a spatiotemporal attention mechanism, which enhances the long-term prediction capabilities of time series models and dynamic spatial modeling. It is applicable to multivariate InSAR spatiotemporal deformation prediction problems. The findings include a list of 72 RTSs compiled based on derived InSAR deformation maps and Sentinel-2 optical images, of which 64 have an average deformation rate exceeding 10 mm/year, indicating signs of permafrost degradation. The density distribution of the displacement maps predicted by the Spacetimeformer model aligned well with the InSAR deformation maps obtained from the small baseline subset (SBAS) method, with the overall prediction deviation controlled within 20 mm. In addition, the point-scale prediction results were compared with LSTM and transformer models. This study indicates that the Spacetimeformer network achieved good results in predicting the deformation of RTSs, with a root mean square error of 1.249 mm. The Spacetimeformer method for deformation prediction with the spacetime mechanism presented in this study can serve as a general framework for multivariate deformation prediction based on InSAR results. It can also quantitatively assess the spatial deformation characteristics and deformation trends of RTSs. Full article

In order to study the root–soil composite system shear characteristics under the action of freeze–thaw cycles in the permafrost regions along the Qinghai–Tibet Highway (QTH) from the Beiluhe–Tuotuohe (B-T) section, the slopes in the permafrost regions along the QTH from the B-T section were selected as the object of the study. The direct shear test of root–soil composite systems under different amounts of freeze–thaw (F-T) cycles and gray correlations were used to analyze the correlation between the number of F-T cycles, water content, root content, and the soil shear strength index. The results show that the cohesion of the soil in the area after F-T cycles exhibits a significant stepwise decrease with an increase in F-T cycles, which can be divided into three stages: the instantaneous stage (a decrease of 46.73–56.42%), the gradual stage (a decrease of 14.80–25.55%), and the stabilization stage (a decrease of 0.61–2.99%). The internal friction angle did not exhibit a regular change. The root–soil composite system showed significant enhancement of soil cohesion compared with soil without roots, with a root content of 0.03 g/cm³ having the most significant effect on soil cohesion (increasing amplitude 65.20–16.82%). With an increase in the number of the F-T cycles, while the water content is greater than 15.0%, the greater the water content of the soil, the smaller its cohesion becomes. Through gray correlation analysis, it was found that the correlation between the number of F-T cycles, water content, root content, and soil cohesion after F-T cycles were 0.63, 0.72, and 0.66, respectively, indicating that water content had the most significant impact on soil cohesion after F-T cycles. The results of this study provide theoretical support for further understanding the variation law of the shear strength of root–soil composite systems in permafrost regions under F-T cycles and the influencing factors of plant roots to enhance soil shear strength under F-T cycles, as well as for the scientific and effective prevention and control of retrogressive thaw slump in the study area, the QTH stretches across the region. Full article

This paper assesses trending AI foundation models, especially emerging computer vision foundation models and their performance in natural landscape feature segmentation. While the term foundation model has quickly garnered interest from the geospatial domain, its definition remains vague. Hence, this paper will first introduce AI foundation models and their defining characteristics. Built upon the tremendous success achieved by Large Language Models (LLMs) as the foundation models for language tasks, this paper discusses the challenges of building foundation models for geospatial artificial intelligence (GeoAI) vision tasks. To evaluate the performance of large AI vision models, especially Meta’s Segment Anything Model (SAM), we implemented different instance segmentation pipelines that minimize the changes to SAM to leverage its power as a foundation model. A series of prompt strategies were developed to test SAM’s performance regarding its theoretical upper bound of predictive accuracy, zero-shot performance, and domain adaptability through fine-tuning. The analysis used two permafrost feature datasets, ice-wedge polygons and retrogressive thaw slumps because (1) these landform features are more challenging to segment than man-made features due to their complicated formation mechanisms, diverse forms, and vague boundaries; (2) their presence and changes are important indicators for Arctic warming and climate change. The results show that although promising, SAM still has room for improvement to support AI-augmented terrain mapping. The spatial and domain generalizability of this finding is further validated using a more general dataset EuroCrops for agricultural field mapping. Finally, we discuss future research directions that strengthen SAM’s applicability in challenging geospatial domains. Full article

Retrogressive thaw slumps (RTSs) are becoming more common on the Qinghai-Tibet Plateau as permafrost thaws, but the hydraulic properties of thaw slumps have not been extensively studied. To fill this knowledge gap, we used the “space-for-time substitution method” to differentiate three stages of RTSs: original grassland, collapsing, and collapsed. Our study included on-site investigations, measurements in the laboratory, and measured and simulated analyses of soil water retention curves and estimated hydrological properties. Our findings show that the measurements and simulated analyses of soil water retention were highly consistent across RTSs, indicating the accuracy of the Van Genuchten model in reproducing soil hydraulic parameters for different stages of RTSs. The original grassland stage had the highest soil water retention and content due to its high soil organic carbon (SOC) content and fine-textured micropores. In contrast, the collapsed stage had higher soil water retention and content compared to the collapsing stage, primarily due to increased proportions of soil micropores, SOC content, and lower bulk density (BD). From original grassland stage to collapsed stage, there were significant changes on the structure of each RTS site, which resulted in a decrease in SOC content and an increase in BD in general. However, the absence of soil structure and compaction led to the subsequent accumulation of organic matter, increasing SOC content. Changes in field capacity, permanent wilting point, and soil micropore distribution aligned with variations in SOC content and water content. These findings highlight the importance of managing SOC content and water content to mitigate the adverse effects of freeze-thaw cycles on soil structure and stability at different thaw collapse stages of RTSs. Effective management strategies may include incorporating organic matter, reducing soil compaction, and maintaining optimal water content. Further research is needed to determine the most suitable management practices for different soil types and environmental conditions. Full article

The Arctic zone of West Siberia (Yamal and Gydan peninsulas) is an area with continuous permafrost and tabular ground ice close to the surface, active thermodenudation, and related landforms: retrogressive thaw slumps (RTS); in Russian referred to as thermocirques (TC). The dimensions of most TCs have not been determined so far. We use Sentinel 2 imagery to measure each TC area ranging from 0.55 to 38 ha with a median of 2.5 ha. Around 95% of TCs have an area of less than 10 ha. The largest areas are gained due to the merging of several neighboring TCs. The ArcticDEM is used to determine TC edge elevation and slope angle. In general, the Median TC of the Yamal peninsula has an area of 1.8 ha, an elevation of the edge of 17.7 m, and a slope angle of 2.5°. The Median TC of the Gydan peninsula has an area of 2.6 ha, elevation of the edge of 29.4 m, and slope angle of 3°. TCs of the Gydan peninsula occupy higher positions and slightly steeper slopes compared to TCs of the Yamal peninsula. The ranges of the median and the largest TC areas are consistent with the reported RTS dimensions in North America. Full article

Ice–rich Pleistocene permafrost deposits (Yedoma) store large amounts of nitrogen (N) and are susceptible to rapid thaw. In this study, we assess whether eroding Yedoma deposits are potential sources of N and gaseous carbon (C) losses. Therefore, we determined aerobic net ammonification and nitrification, as well as anaerobic production of nitrous oxide (N₂O), carbon dioxide (CO₂), and methane (CH₄) in laboratory incubations. Samples were collected from non-vegetated and revegetated slump floor (SF) and thaw mound (TM) soils of a retrogressive thaw slump in the Lena River Delta of Eastern Siberia. We found high nitrate concentrations (up to 110 µg N (g DW)⁻¹) within the growing season, a faster transformation of organic N to nitrate, and high N₂O production (up to 217 ng N₂O-N (g DW)⁻¹ day⁻¹) in revegetated thaw mounds. The slump floor was low in nitrate and did not produce N₂O under anaerobic conditions, but produced the most CO₂ (up to 7 µg CO₂-C (g DW)⁻¹ day⁻¹) and CH₄ (up to 65 ng CH₄-C (g DW)⁻¹ day⁻¹). Nitrate additions showed that denitrification was substrate limited in the slump floor. Nitrate limitation was rather caused by field conditions (moisture, pH) than by microbial functional limitation since nitrification rates were positive under laboratory conditions. Our results emphasize the relevance of considering landscape processes, geomorphology, and soil origin in order to identify hotspots of high N availability, as well as C and N losses. High N availability is likely to have an impact on carbon cycling, but to what extent needs further investigation. Full article

Ice-rich permafrost in the Qinghai–Tibet Plateau (QTP), China, is becoming susceptible to thermokarst landforms, and the most dramatic among these terrain-altering landforms is retrogressive thaw slump (RTS). Concurrently, RTS development can in turn affect the eco-environment, and especially soil erosion and carbon emission, during their evolution. However, there are still a lack of quantitative methods and comprehensive studies on the deformation and volumetric change in RTS. The purpose of this study is to quantitatively assess the RTS evolution through a novel and feasible simulation framework of the GPU-based discrete element method (DEM) coupled with the finite difference method (FDM). Additionally, the simulation results were calibrated using the time series observation results from September 2021 to August 2022, using the combined methods of terrestrial laser scanning (TLS) and unmanned aerial vehicle (UAV). The results reveal that, over this time, thaw slump mobilized a total volume of 1335 m³ and approximately 1050 m³ moved to a displaced area. Additionally, the estimated soil erosion was about 211 m³. Meanwhile, the corresponding maximum ground subsidence and headwall retrogression were 1.9 m and 3.2 m, respectively. We also found that the amount of mass wasting in RTS development is highly related to the ground ice content. When the volumetric ice content exceeds 10%, there will be obvious mass wasting in the thaw slump development area. Furthermore, this work proposed that the coupled DEM-FDM method and field survey method of TLS-UAV can provide an effective pathway to simulate thaw-induced slope failure problems and complement the research limitations of small-scale RTSs using remote sensing methods. The results are meaningful for assessing the eco-environmental impacts on the QTP. Full article

Retrogressive thaw slumps (RTS) are considered one of the most dynamic permafrost disturbance features in the Arctic. Sub-meter resolution multispectral imagery acquired by very high spatial resolution (VHSR) commercial satellite sensors offer unique capacities in capturing the morphological dynamics of RTSs. The central goal of this study is to develop a deep learning convolutional neural net (CNN) model (a UNet-based workflow) to automatically detect and characterize RTSs from VHSR imagery. We aimed to understand: (1) the optimal combination of input image tile size (array size) and the CNN network input size (resizing factor/spatial resolution) and (2) the interoperability of the trained UNet models across heterogeneous study sites based on a limited set of training samples. Hand annotation of RTS samples, CNN model training and testing, and interoperability analyses were based on two study areas from high-Arctic Canada: (1) Banks Island and (2) Axel Heiberg Island and Ellesmere Island. Our experimental results revealed the potential impact of image tile size and the resizing factor on the detection accuracies of the UNet model. The results from the model transferability analysis elucidate the effects on the UNet model due the variability (e.g., shape, color, and texture) associated with the RTS training samples. Overall, study findings highlight several key factors that we should consider when operationalizing CNN-based RTS mapping over large geographical extents. Full article

Deep learning has been used for mapping retrogressive thaw slumps and other periglacial landforms but its application is still limited to local study areas. To understand the accuracy, efficiency, and transferability of a deep learning model (i.e., DeepLabv3+) when applied to large areas or multiple regions, we conducted several experiments using training data from three different regions across the Canadian Arctic. To overcome the main challenge of transferability, we used a generative adversarial network (GAN) called CycleGAN to produce new training data in an attempt to improve transferability. The results show that (1) data augmentation can improve the accuracy of the deep learning model but does not guarantee transferability, (2) it is necessary to choose a good combination of hyper-parameters (e.g., backbones and learning rate) to achieve an optimal trade-off between accuracy and efficiency, and (3) a GAN can significantly improve the transferability if the variation between source and target is dominated by color or general texture. Our results suggest that future mapping of retrogressive thaw slumps should prioritize the collection of training data from regions where a GAN cannot improve the transferability. Full article

In a warming Arctic, permafrost-related disturbances, such as retrogressive thaw slumps (RTS), are becoming more abundant and dynamic, with serious implications for permafrost stability and bio-geochemical cycles on local to regional scales. Despite recent advances in the field of earth observation, many of these have remained undetected as RTS are highly dynamic, small, and scattered across the remote permafrost region. Here, we assessed the potential strengths and limitations of using deep learning for the automatic segmentation of RTS using PlanetScope satellite imagery, ArcticDEM and auxiliary datasets. We analyzed the transferability and potential for pan-Arctic upscaling and regional cross-validation, with independent training and validation regions, in six different thaw slump-affected regions in Canada and Russia. We further tested state-of-the-art model architectures (UNet, UNet++, DeepLabv3) and encoder networks to find optimal model configurations for potential upscaling to continental scales. The best deep learning models achieved mixed results from good to very good agreement in four of the six regions (maxIoU: 0.39 to 0.58; Lena River, Horton Delta, Herschel Island, Kolguev Island), while they failed in two regions (Banks Island, Tuktoyaktuk). Of the tested architectures, UNet++ performed the best. The large variance in regional performance highlights the requirement for a sufficient quantity, quality and spatial variability in the training data used for segmenting RTS across diverse permafrost landscapes, in varying environmental conditions. With our highly automated and configurable workflow, we see great potential for the transfer to active RTS clusters (e.g., Peel Plateau) and upscaling to much larger regions. Full article