Search Results (18)

Ice wedges, which are ubiquitous in permafrost areas, play a significant role in the evolution of permafrost landscapes, influencing the topography and hydrology of these regions. In this paper, we combine a detailed multi-generational, interdisciplinary, and international literature review along with our own field experiences to explore the development of low-centered ice-wedge polygons and their orthogonal networks. Low-centered polygons, a type of ice-wedge polygonal ground characterized by elevated rims and lowered wet central basins, are critical indicators of permafrost conditions. The formation of these features has been subject to numerous inconsistencies and debates since their initial description in the 1800s. The development of elevated rims is attributed to different processes, such as soil bulging due to ice-wedge growth, differential frost heave, and the accumulation of vegetation and peat. The transition of low-centered polygons to flat-centered, driven by processes like peat accumulation, aggradational ice formation, and frost heave in polygon centers, has been generally overlooked. Low-centered polygons occur in deltas, on floodplains, and in drained-lake basins. There, they are often arranged in orthogonal networks that comprise a complex system. The prevailing explanation of their formation does not match with several field studies that practically remain unnoticed or ignored. By analyzing controversial subjects, such as the degradational or aggradational nature of low-centered polygons and the formation of orthogonal ice-wedge networks, this paper aims to clarify misconceptions and present a cohesive overview of lowland terrain ice-wedge dynamics. The findings emphasize the critical role of ice wedges in shaping Arctic permafrost landscapes and their vulnerability to ongoing climatic and landscape changes. Full article

Fine-scale maps of ground ice and related surface features are critical for permafrost-related modelling and management. However, such maps are lacking across almost the entire Arctic. Machine learning provides the potential to automate regional fine-scale ground ice mapping using remote sensing and topographic data. Here, we evaluate the predictive skill of XGBoost models for identifying (1) ice wedge and (2) top-5m visible ground ice in the Tuktoyaktuk Coastlands. We find high predictive skill for ice wedge occurrence (ROC AUC = 0.95, macro F1 = 0.80), with the most important predictors being slope, distance to the coast, and probability of depression. The model accurately predicted regional and local trends in ice wedge occurrence, with an increase in ice wedge polygon (IWP) probability towards the coast and in poorly drained depressions. The model also captured IWP in well-drained uplands of the study area, including locations with poorly visible troughs not contained in the training data. Spatial transferability analyses highlight the regional variability of ice wedge probability, reflecting contrasting climatic and surface conditions. Conversely, the low predictive skill for visible ground ice (ROC AUC = 0.67, macro F1 = 0.53) is attributed to limitations in training data and weak associations with the remotely sensed predictors. The varying predictive accuracy highlights the importance of high-quality reference data and site-specific conditions for improving ground ice studies with data-driven modelling from remote sensing observations. 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

As one of the best indicators of the periglacial environment, ice-wedge polygons (IWPs) are important for arctic landscapes, hydrology, engineering, and ecosystems. Thus, a better understanding of the spatiotemporal dynamics and evolution of IWPs is key to evaluating the hydrothermal state and carbon budgets of the arctic permafrost environment. In this paper, the dynamics of ground surface deformation (GSD) in IWP zones (2018–2019) and their influencing factors over the last 20 years in Saskylakh, northwestern Yakutia, Russia were investigated using the Interferometric Synthetic Aperture Radar (InSAR) and Google Earth Engine (GEE). The results show an annual ground surface deformation rate (AGSDR) in Saskylakh at −49.73 to 45.97 mm/a during the period from 1 June 2018 to 3 May 2019. All the selected GSD regions indicate that the relationship between GSD and land surface temperature (LST) is positive (upheaving) for regions with larger AGSDR, and negative (subsidence) for regions with lower AGSDR. The most drastic deformation was observed at the Aeroport regions with GSDs rates of −37.06 mm/a at tower and 35.45 mm/a at runway. The GSDs are negatively correlated with the LST of most low-centered polygons (LCPs) and high-centered polygons (HCPs). Specifically, the higher the vegetation cover, the higher the LST and the thicker the active layer. An evident permafrost degradation has been observed in Saskylakh as reflected in higher ground temperatures, lusher vegetation, greater active layer thickness, and fluctuant numbers and areal extents of thermokarst lakes and ponds. Full article

Ground subsidence and erosion processes caused by permafrost thaw pose a high risk to infrastructure in the Arctic. Climate warming is increasingly accelerating the thawing of permafrost, emphasizing the need for thorough monitoring to detect damages and hazards at an early stage. The use of unoccupied aerial vehicles (UAVs) allows a fast and uncomplicated analysis of sub-meter changes across larger areas compared to manual surveys in the field. In our study, we investigated the potential of photogrammetry products derived from imagery acquired with off-the-shelf UAVs in order to provide a low-cost assessment of the risks of permafrost degradation along critical infrastructure. We tested a minimal drone setup without ground control points to derive high-resolution 3D point clouds via structure from motion (SfM) at a site affected by thermal erosion along the Dalton Highway on the North Slope of Alaska. For the sub-meter change analysis, we used a multiscale point cloud comparison which we improved by applying (i) denoising filters and (ii) alignment procedures to correct for horizontal and vertical offsets. Our results show a successful reduction in outliers and a thorough correction of the horizontal and vertical point cloud offset by a factor of 6 and 10, respectively. In a defined point cloud subset of an erosion feature, we derive a median land surface displacement of $-0.35$ m from 2018 to 2019. Projecting the development of the erosion feature, we observe an expansion to NNE, following the ice-wedge polygon network. With a land surface displacement of $-0.35$ m and an alignment root mean square error of $0.99$ m, we find our workflow is best suitable for detecting and quantifying rapid land surface changes. For a future improvement of the workflow, we recommend using alternate flight patterns and an enhancement of the point cloud comparison algorithm. Full article

In response to increasing Arctic temperatures, ice-rich permafrost landscapes are undergoing rapid changes. In permafrost lowlands, polygonal ice wedges are especially prone to degradation. Melting of ice wedges results in deepening troughs and the transition from low-centered to high-centered ice-wedge polygons. This process has important implications for surface hydrology, as the connectivity of such troughs determines the rate of drainage for these lowland landscapes. In this study, we present a comprehensive, modular, and highly automated workflow to extract, to represent, and to analyze remotely sensed ice-wedge polygonal trough networks as a graph (i.e., network structure). With computer vision methods, we efficiently extract the trough locations as well as their geomorphometric information on trough depth and width from high-resolution digital elevation models and link these data within the graph. Further, we present and discuss the benefits of graph analysis algorithms for characterizing the erosional development of such thaw-affected landscapes. Based on our graph analysis, we show how thaw subsidence has progressed between 2009 and 2019 following burning at the Anaktuvuk River fire scar in northern Alaska, USA. We observed a considerable increase in the number of discernible troughs within the study area, while simultaneously the number of disconnected networks decreased from 54 small networks in 2009 to only six considerably larger disconnected networks in 2019. On average, the width of the troughs has increased by 13.86%, while the average depth has slightly decreased by 10.31%. Overall, our new automated approach allows for monitoring ice-wedge dynamics in unprecedented spatial detail, while simultaneously reducing the data to quantifiable geometric measures and spatial relationships. Full article

Land-atmosphere carbon exchange is known to be extremely heterogeneous in arctic ice-wedge polygonal tundra regions. In this study, a Kalman filter-based method was developed to estimate the spatio-temporal dynamics of daytime average net ecosystem exchange (NEEday) at 0.5-m resolution over a 550 m by 700 m study site. We integrated multi-scale, multi-type datasets, including normalized difference vegetation indices (NDVIs) obtained from a novel automated mobile sensor system (or tram system) and a greenness index map obtained from airborne imagery. We took advantage of the significant correlations between NDVI and NEEday identified based on flux chamber measurements. The weighted average of the estimated NEEday within the flux-tower footprint agreed with the flux tower data in term of its seasonal dynamics. We then evaluated the spatial variability of the growing season average NEEday, as a function of polygon geomorphic classes; i.e., the combination of polygon types—which are known to present different degradation stages associated with permafrost thaw—and microtopographic features (i.e., troughs, centers and rims). Our study suggests the importance of considering microtopographic features and their spatial coverage in computing spatially aggregated carbon exchange. Full article

Very high spatial resolution commercial satellite imagery can inform observation, mapping, and documentation of micro-topographic transitions across large tundra regions. The bridging of fine-scale field studies with pan-Arctic system assessments has until now been constrained by a lack of overlap in spatial resolution and geographical coverage. This likely introduced biases in climate impacts on, and feedback from the Arctic region to the global climate system. The central objective of this exploratory study is to develop an object-based image analysis workflow to automatically extract ice-wedge polygon troughs from very high spatial resolution commercial satellite imagery. We employed a systematic experiment to understand the degree of interoperability of knowledge-based workflows across distinct tundra vegetation units—sedge tundra and tussock tundra—focusing on the same semantic class. In our multi-scale trough modelling workflow, we coupled mathematical morphological filtering with a segmentation process to enhance the quality of image object candidates and classification accuracies. Employment of the master ruleset on sedge tundra reported classification accuracies of correctness of 0.99, completeness of 0.87, and F1 score of 0.92. When the master ruleset was applied to tussock tundra without any adaptations, classification accuracies remained promising while reporting correctness of 0.87, completeness of 0.77, and an F1 score of 0.81. Overall, results suggest that the object-based image analysis-based trough modelling workflow exhibits substantial interoperability across the terrain while producing promising classification accuracies. From an Arctic earth science perspective, the mapped troughs combined with the ArcticDEM can allow hydrological assessments of lateral connectivity of the rapidly changing Arctic tundra landscape, and repeated mapping can allow us to track fine-scale changes across large regions and that has potentially major implications on larger riverine systems. Full article

Spatiotemporal variation in ground-surface displacement caused by ground freeze–thaw and thermokarst is critical information to understand changes in the permafrost ecosystem. Measurement of ground displacement, especially in the disturbed ground underlain by ice-rich permafrost, is important to estimate the rate of permafrost and carbon loss. We conducted high-precision global navigation satellite system (GNSS) positioning surveys to measure the surface displacements of tundra in northern Alaska, together with maximum thaw depth (TD) and surface moisture measurements from 2017 to 2019. The measurements were performed along two to three 60–200 m transects per site with 1–5 m intervals at the three areas. The average seasonal thaw settlement (STS) at intact tundra sites ranged 5.8–14.3 cm with a standard deviation range of 2.1–3.3 cm. At the disturbed locations, averages and variations in STS and the maximum thaw depth were largest in all observed years and among all sites. The largest seasonal and interannual subsidence (44 and 56 cm/year, respectively) were recorded at points near troughs of degraded ice-wedge polygons or thermokarst lakes. Weak or moderate correlation between STS and TD found at the intact sites became obscure as the thermokarst disturbance progressed, leading to higher uncertainty in the prediction of TD from STS. Full article

We developed a high-throughput mapping workflow, which centers on deep learning (DL) convolutional neural network (CNN) algorithms on high-performance distributed computing resources, to automatically characterize ice-wedge polygons (IWPs) from sub-meter resolution commercial satellite imagery. We applied a region-based CNN object instance segmentation algorithm, namely the Mask R-CNN, to automatically detect and classify IWPs in North Slope of Alaska. The central goal of our study was to systematically expound the DLCNN model interoperability across varying tundra types (sedge, tussock sedge, and non-tussock sedge) and image scene complexities to refine the understanding of opportunities and challenges for regional-scale mapping applications. We corroborated quantitative error statistics along with detailed visual inspections to gauge the IWP detection accuracies. We found promising model performances (detection accuracies: 89% to 96% and classification accuracies: 94% to 97%) for all candidate image scenes with varying tundra types. The mapping workflow discerned the IWPs by exhibiting low absolute mean relative error (AMRE) values (0.17–0.23). Results further suggest the importance of increasing the variability of training samples when practicing transfer-learning strategy to map IWPs across heterogeneous tundra cover types. Overall, our findings demonstrate the robust performances of IWPs mapping workflow in multiple tundra landscapes. Full article

As ice wedge degradation and the inundation of polygonal troughs become increasingly common processes across the Arctic, lateral export of water from polygonal soils may represent an important mechanism for the mobilization of dissolved organic carbon and other solutes. However, drainage from ice wedge polygons is poorly understood. We constructed a model which uses cross-sectional flow nets to define flow paths of meltwater through the active layer of an inundated low-centered polygon towards the trough. The model includes the effects of evaporation and simulates the depletion of ponded water in the polygon center during the thaw season. In most simulations, we discovered a strong hydrodynamic edge effect: only a small fraction of the polygon volume near the rim area is flushed by the drainage at relatively high velocities, suggesting that nearly all advective transport of solutes, heat, and soil particles is confined to this zone. Estimates of characteristic drainage times from the polygon center are consistent with published field observations. Full article

Deep learning (DL) convolutional neural networks (CNNs) have been rapidly adapted in very high spatial resolution (VHSR) satellite image analysis. DLCNN-based computer visions (CV) applications primarily aim for everyday object detection from standard red, green, blue (RGB) imagery, while earth science remote sensing applications focus on geo object detection and classification from multispectral (MS) imagery. MS imagery includes RGB and narrow spectral channels from near- and/or middle-infrared regions of reflectance spectra. The central objective of this exploratory study is to understand to what degree MS band statistics govern DLCNN model predictions. We scaffold our analysis on a case study that uses Arctic tundra permafrost landform features called ice-wedge polygons (IWPs) as candidate geo objects. We choose Mask RCNN as the DLCNN architecture to detect IWPs from eight-band Worldview-02 VHSR satellite imagery. A systematic experiment was designed to understand the impact on choosing the optimal three-band combination in model prediction. We tasked five cohorts of three-band combinations coupled with statistical measures to gauge the spectral variability of input MS bands. The candidate scenes produced high model detection accuracies for the F1 score, ranging between 0.89 to 0.95, for two different band combinations (coastal blue, blue, green (1,2,3) and green, yellow, red (3,4,5)). The mapping workflow discerned the IWPs by exhibiting low random and systematic error in the order of 0.17–0.19 and 0.20–0.21, respectively, for band combinations (1,2,3). Results suggest that the prediction accuracy of the Mask-RCNN model is significantly influenced by the input MS bands. Overall, our findings accentuate the importance of considering the image statistics of input MS bands and careful selection of optimal bands for DLCNN predictions when DLCNN architectures are restricted to three spectral channels. Full article

State-of-the-art deep learning technology has been successfully applied to relatively small selected areas of very high spatial resolution (0.15 and 0.25 m) optical aerial imagery acquired by a fixed-wing aircraft to automatically characterize ice-wedge polygons (IWPs) in the Arctic tundra. However, any mapping of IWPs at regional to continental scales requires images acquired on different sensor platforms (particularly satellite) and a refined understanding of the performance stability of the method across sensor platforms through reliable evaluation assessments. In this study, we examined the transferability of a deep learning Mask Region-Based Convolutional Neural Network (R-CNN) model for mapping IWPs in satellite remote sensing imagery (~0.5 m) covering 272 km² and unmanned aerial vehicle (UAV) (0.02 m) imagery covering 0.32 km². Multi-spectral images were obtained from the WorldView-2 satellite sensor and pan-sharpened to ~0.5 m, and a 20 mp CMOS sensor camera onboard a UAV, respectively. The training dataset included 25,489 and 6022 manually delineated IWPs from satellite and fixed-wing aircraft aerial imagery near the Arctic Coastal Plain, northern Alaska. Quantitative assessments showed that individual IWPs were correctly detected at up to 72% and 70%, and delineated at up to 73% and 68% F1 score accuracy levels for satellite and UAV images, respectively. Expert-based qualitative assessments showed that IWPs were correctly detected at good (40–60%) and excellent (80–100%) accuracy levels for satellite and UAV images, respectively, and delineated at excellent (80–100%) level for both images. We found that (1) regardless of spatial resolution and spectral bands, the deep learning Mask R-CNN model effectively mapped IWPs in both remote sensing satellite and UAV images; (2) the model achieved a better accuracy in detection with finer image resolution, such as UAV imagery, yet a better accuracy in delineation with coarser image resolution, such as satellite imagery; (3) increasing the number of training data with different resolutions between the training and actual application imagery does not necessarily result in better performance of the Mask R-CNN in IWPs mapping; (4) and overall, the model underestimates the total number of IWPs particularly in terms of disjoint/incomplete IWPs. Full article

Modern degradation of Arctic permafrost promotes changes in tundra landscapes and leads to degradation of ice wedge polygons, which are the most widespread landforms of Arctic wetlands. Status assessment of polygon degradation is important for various environmental studies. We have applied the geographic information systems’ (GIS) analysis of data from unmanned aerial vehicles (UAV) to accurately assess the status of ice wedge polygon degradation on Samoylov Island. We used several modern models of polygon degradation for revealing polygon types, which obviously correspond to different stages of degradation. Manual methods of mapping and a high spatial resolution of used UAV data allowed for a high degree of accuracy in the identification of all land units. The study revealed the following: 41.79% of the first terrace surface was composed of non-degraded polygonal tundra; 18.37% was composed of polygons, which had signs of thermokarst activity and corresponded to various stages of degradation in the models; and 39.84% was composed of collapsed polygons, slopes, valleys, and water bodies, excluding ponds of individual polygons. This study characterizes the current status of polygonal tundra degradation of the first terrace surface on Samoylov Island. Our assessment reflects the landscape condition of the first terrace surface of Samoylov Island, which is the typical island of the southern part of the Lena Delta. Moreover, the study illustrates the potential of UAV data GIS analysis for highly accurate investigations of Arctic landscape changes. Full article

Eastern Siberia is characterized by widespread permafrost thawing and subsequent thermokarst development. Estimation of the impacts of the predicted rise in precipitation and air temperatures under climate change requires quantitative knowledge about the spatial distribution of thermokarst development. In the last few years, unmanned aerial systems (UAS) and structure-from-motion multi-view stereo (SfM-MVS) photogrammetry attracted a tremendous amount of interest for acquiring high-definition topographic data. This study detected characteristics of thermokarst landforms using UAS and SfM-MVS photogrammetry at a disused airfield (3.0 ha) and for arable land that was previously used for farming (6.3 ha) in the Churapcha area, located on the right bank of the Lena River in central Yakutia. Orthorectified photographs and digital terrain models with spatial resolutions of 4.0 cm and 8.0 cm, respectively, were obtained for this study. At the disused airfield site and the abandoned arable land, 174 and 867 high-centered polygons that developed after the 1990s were detected, respectively. The data showed that the average diameter and average area of the polygons at the disused airfield site were 11.6 m and 111.2 m², respectively, while those of the polygons in the abandoned arable land were 7.4 m and 46.8 m², respectively. The abandoned arable land was characterized by smaller polygons and a higher polygon density. The differences in polygon size for the abandoned arable land and the disused airfield site indicate a difference in the ice wedge distributions and thermokarst developments. The subsidence rate was estimated as 2.1 cm/year for the disused airfield site and 3.9 cm/year for the abandoned arable land. Full article