Search Results (31)

Timberline marks the transitions from continuous forests to sparse forests and tundra landscapes. As the spatial distribution and dynamics of timberline are closely associated with regional energy and carbon balance, mapping timberline is important to a wide range of environmental and ecological studies. However, current timberline delineation approaches remain under-developed. We proposed an automatic timberline delineation method based on a seeded region-growing segmentation technique and satellite-derived products of tree fractional cover. We applied our approach to the West Siberian Plain and Alaska treeline regions as defined by the Circumpolar Arctic Vegetation Map. The results demonstrate the effectiveness of the proposed method for the accurate delineation of the timberlines that spatially align well with very-high-resolution satellite images. Based on the delineated timberlines, we find regional-scale tree encroachment to be not as substantial as previously reported. The proposed approach can be applied to understanding climate-induced forest responses and inform forest management practices. Full article

We explore a new approach for the parsimonious, generalizable, efficient, and potentially automatable characterization of spectral diversity of sparse targets in spectroscopic imagery. The approach focuses on pixels which are not well modeled by linear subpixel mixing of the Substrate, Vegetation and Dark (S, V, and D) endmember spectra which dominate spectral variance for most of Earth’s land surface. We illustrate the approach using AVIRIS-3 imagery of anthropogenic surfaces (primarily hydrocarbon extraction infrastructure) embedded in a background of Arctic tundra near Prudhoe Bay, Alaska. Computational experiments further explore sensitivity to spatial and spectral resolution. Analysis involves two stages: first, computing the mixture residual of a generalized linear spectral mixture model; and second, nonlinear dimensionality reduction via manifold learning. Anthropogenic targets and lakeshore sediments are successfully isolated from the Arctic tundra background. Dependence on spatial resolution is observed, with substantial degradation of manifold topology as images are blurred from 5 m native ground sampling distance to simulated 30 m ground projected instantaneous field of view of a hypothetical spaceborne sensor. Degrading spectral resolution to mimicking the Sentinel-2A MultiSpectral Imager (MSI) also results in loss of information but is less severe than spatial blurring. These results inform spectroscopic characterization of sparse targets using spectroscopic images of varying spatial and spectral resolution. Full article

Climate warming significantly impacts Arctic vegetation, yet its future role as a carbon sink or source is unclear. We analyzed vegetation growth and carbon exchange in Alaska’s tundra and needle leaf forests using the LPJ-GUESS model. The accuracy of the model is verified using linear regression of the measured data from 2004 to 2008, and the results are significantly correlated, which proves that the model is reliable, with R² values of 0.51 and 0.46, respectively, for net ecosystem carbon exchange (NEE) at the tundra and needle leaf forest sites, and RMSE values of 22.85 and 23.40 gC/m²/yr for the tundra and needle forest sites, respectively. For the gross primary production (GPP), the R² values were 0.66 and 0.85, and the RMSE values were 39.25 and 43.75 gC/m²/yr at the tundra and needle leaf forest sites, respectively. We simulated vegetation carbon exchanges for 1992–2014 and projected future exchanges for 2020–2100 using climate variables. Under SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios, GPP values increase with higher emissions, while the NEE showed great fluctuations without significant differences among the three pathways. Our results showed although climate warming can benefit vegetation growth, net carbon assimilation by vegetation may not increase accordingly in the future. Full article

Plot-level photography is an attractive time-saving alternative to field measurements for vegetation monitoring. However, widespread adoption of this technique relies on efficient workflows for post-processing images and the accuracy of the resulting products. Here, we estimated relative vegetation cover using both traditional field sampling methods (point frame) and semi-automated classification of photographs (plot-level photography) across thirty 1 m² plots near Utqiaġvik, Alaska, from 2012 to 2021. Geographic object-based image analysis (GEOBIA) was applied to generate objects based on the three spectral bands (red, green, and blue) of the images. Five machine learning algorithms were then applied to classify the objects into vegetation groups, and random forest performed best (60.5% overall accuracy). Objects were reliably classified into the following classes: bryophytes, forbs, graminoids, litter, shadows, and standing dead. Deciduous shrubs and lichens were not reliably classified. Multinomial regression models were used to gauge if the cover estimates from plot-level photography could accurately predict the cover estimates from the point frame across space or time. Plot-level photography yielded useful estimates of vegetation cover for graminoids. However, the predictive performance varied both by vegetation class and whether it was being used to predict cover in new locations or change over time in previously sampled plots. These results suggest that plot-level photography may maximize the efficient use of time, funding, and available technology to monitor vegetation cover in the Arctic, but the accuracy of current semi-automated image analysis is not sufficient to detect small changes in cover. Full article

Few fires are known to have burned the tundra of the Arctic Slope north of the Brooks Range in Alaska, USA. A total of 90 fires between 1969 and 2022 are known. Because fire has been rare, old burns can be detected by the traces of thermokarst and distinct vegetation they leave in otherwise uniform tundra, which are visible in aerial photograph archives. Several prehistoric tundra burns have been found in this way. Detection of tundra fires in this sparsely populated and remote area has been historically inconsistent and opportunistic, relying on reports by aircraft pilots. Fire reports have been logged into an administrative database which, out of necessity, has been used to scientifically evaluate changes in the fire regime. To improve the consistency of the record, we completed a systematic search of Landsat Collection 2 for the Brooks Range Foothills ecoregion over the period 1972–2022. We found 57 unrecorded tundra burns, about 41% of the total, which now numbers 138. Only 15% and 33% of all fires appear in MODIS and VIIRS satellite-borne thermal anomaly products, respectively. The fire frequency in the first 37 years of the record is 0.89 y⁻¹ for natural ignitions that spread ≥10 ha. Frequency in the last 13 years is 2.5 y⁻¹, indicating a nearly three-fold increase in fire frequency. Full article

Permafrost tundra contains more than twice as much carbon as is currently in the atmosphere, and it is warming six times as fast as the global mean. Tundra lakes dynamics is a robust indicator of global climate processes, and is still not well understood. Satellite data, particularly, from synthetic aperture radar (SAR) is a suitable tool for tundra lakes recognition and monitoring of their changes. However, manual analysis of lake boundaries can be slow and inefficient; therefore, reliable automated algorithms are required. To address this issue, we propose a two-stage approach, comprising instance deep-learning-based segmentation by U-Net, followed by semantic segmentation based on a watershed algorithm for separating touching and overlapping lakes. Implementation of this concept is essential for accurate sizes and shapes estimation of an individual lake. Here, we evaluated the performance of the proposed approach on lakes, manually extracted from tens of C-band SAR images from Sentinel-1, which were collected in the Yamal Peninsula and Alaska areas in the summer months of 2015–2022. An accuracy of 0.73, in terms of the Jaccard similarity index, was achieved. The lake’s perimeter, area and fractal sizes were estimated, based on the algorithm framework output from hundreds of SAR images. It was recognized as lognormal distributed. The evaluation of the results indicated the efficiency of the proposed approach for accurate automatic estimation of tundra lake shapes and sizes, and its potential to be used for further studies on tundra lake dynamics, in the context of global climate change, aimed at revealing new factors that could cause the planet to warm or cool. Full article

Global estimates of burned areas, enabled by the wide-open access to the standard data products from the Moderate Resolution Imaging Spectroradiometer (MODIS), are heavily relied on by scientists and managers studying issues related to wildfire occurrence and its worldwide consequences. While these datasets, particularly the MODIS MCD64A1 product, have fundamentally improved our understanding of wildfire regimes at the global scale, their performance may be less reliable in certain regions due to a series of region- or ecosystem-specific challenges. Previous studies have indicated that global burned area products tend to underestimate the extent of the burned area within some parts of the boreal domain. Despite this, global products are still being regularly used by research activities and management efforts in the northern regions, likely due to a lack of understanding of the spatial scale of their Arctic-specific limitations, as well as an absence of more reliable alternative products. In this study, we evaluated the performance of two widely used global burned area products, MCD64A1 and FireCCI51, in the circumpolar boreal forests and tundra between 2001 and 2015. Our two-step evaluation shows that MCD64A1 has high commission and omission errors in mapping burned areas in the boreal forests and tundra regions in North America. The omission error overshadows the commission error, leading to MCD64A1 considerably underestimating burned areas in these high northern latitude domains. Based on our estimation, MCD64A1 missed nearly half the total burned areas in the Alaskan and Canadian boreal forests and the tundra during the 15-year period, amounting to an area (74,768 km²) that is equivalent to the land area of the United States state of South Carolina. While the FireCCI51 product performs much better than MCD64A1 in terms of commission error, we found that it also missed about 40% of burned areas in North America north of 60° N between 2001 and 2015. Our intercomparison of MCD64A1 and FireCCI51 with a regionally adapted MODIS-based Arctic Boreal Burned Area (ABBA) shows that the latter outperforms both MCD64A1 and FireCCI51 by a large margin, particularly in terms of omission error, and thus delivers a considerably more accurate and consistent estimate of fire activity in the high northern latitudes. Considering the fact that boreal forests and tundra represent the largest carbon pool on Earth and that wildfire is the dominant disturbance agent in these ecosystems, our study presents a strong case for regional burned area products like ABBA to be included in future Earth system models as the critical input for understanding wildfires’ impacts on global carbon cycling and energy budget. Full article

Daily Normalized Difference Vegetation Index (NDVI) values from the MODIS Aqua and Terra satellites were compared with on-the-ground camera observations at five locations in northern Alaska. Over half of the spring rise in NDVI was due to the transition from the snow-covered landscape to the snow-free surface prior to the deciduous leaf-out. In the fall after the green season, NDVI fluctuated between an intermediate level representing senesced vegetation and lower values representing clouds and intermittent snow, and then dropped to constant low levels after establishment of the permanent winter snow cover. The NDVI value of snow-free surfaces after fall leaf senescence was estimated from multi-year data using a 90th percentile smoothing spline curve fit to a plot of daily NDVI values vs. ordinal date. This curve typically showed a flat region of intermediate NDVI values in the fall that represent cloud- and snow-free days with senesced vegetation. This “fall plateau” was readily identified in a large systematic sample of MODIS NDVI values across the study area, in typical tundra, shrub, and boreal forest environments. The NDVI level of the fall plateau can be extrapolated to the spring rising leg of the annual NDVI curve to approximate the true start of green season. Full article

The Arctic’s winter water cycle is rapidly changing, with implications for snow moisture sources and transport processes. Stable isotope values (δ¹⁸O, δ²H, d-excess) of the Arctic snowpack have potential to provide proxy records of these processes, yet it is unclear how well the isotope values of individual snowfall events are preserved within snow profiles. Here, we present water isotope data from multiple taiga and tundra snow profiles sampled in Arctic Alaska and Finland, respectively, during winter 2018–2019. We compare the snowpack isotope stratigraphy with meteoric water isotopes (vapor and precipitation) during snowfall days, and combine our measurements with satellite observations and reanalysis data. Our analyses indicate that synoptic-scale atmospheric circulation and regional sea ice coverage are key drivers of the source, amount, and isotopic composition of Arctic snowpacks. We find that the western Arctic tundra snowpack profiles in Alaska preserved the isotope values for the most recent storm; however, post depositional processes modified the remaining isotope profiles. The overall seasonal evolution in the vapor isotope values were better preserved in taiga snow isotope profiles in the eastern Arctic, where there is significantly less wind-driven redistribution than in the open Alaskan tundra. We demonstrate the potential of the seasonal snowpack to provide a useful proxy for Arctic winter-time moisture sources and propose future analyses. 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

Changes in vegetation distribution, structure, and function can modify the canopy properties of terrestrial ecosystems, with potential consequences for regional and global climate feedbacks. In the Arctic, climate is warming twice as fast as compared to the global average (known as ‘Arctic amplification’), likely having stronger impacts on arctic tundra vegetation. In order to quantify these changes and assess their impacts on ecosystem structure and function, methods are needed to accurately characterize the canopy properties of tundra vegetation types. However, commonly used ground-based measurements are limited in spatial and temporal coverage, and differentiating low-lying tundra plant species is challenging with coarse-resolution satellite remote sensing. The collection and processing of multi-sensor data from unoccupied aerial systems (UASs) has the potential to fill the gap between ground-based and satellite observations. To address the critical need for such data in the Arctic, we developed a cost-effective multi-sensor UAS (the ‘Osprey’) using off-the-shelf instrumentation. The Osprey simultaneously produces high-resolution optical, thermal, and structural images, as well as collecting point-based hyperspectral measurements, over vegetation canopies. In this paper, we describe the setup and deployment of the Osprey system in the Arctic to a tundra study site located in the Seward Peninsula, Alaska. We present a case study demonstrating the processing and application of Osprey data products for characterizing the key biophysical properties of tundra vegetation canopies. In this study, plant functional types (PFTs) representative of arctic tundra ecosystems were mapped with an overall accuracy of 87.4%. The Osprey image products identified significant differences in canopy-scale greenness, canopy height, and surface temperature among PFTs, with deciduous low to tall shrubs having the lowest canopy temperatures while non-vascular lichens had the warmest. The analysis of our hyperspectral data showed that variation in the fractional cover of deciduous low to tall shrubs was effectively characterized by Osprey reflectance measurements across the range of visible to near-infrared wavelengths. Therefore, the development and deployment of the Osprey UAS, as a state-of-the-art methodology, has the potential to be widely used for characterizing tundra vegetation composition and canopy properties to improve our understanding of ecosystem dynamics in the Arctic, and to address scale issues between ground-based and airborne/satellite observations. 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

The microtopography associated with ice-wedge polygons governs many aspects of Arctic ecosystem, permafrost, and hydrologic dynamics from local to regional scales owing to the linkages between microtopography and the flow and storage of water, vegetation succession, and permafrost dynamics. Wide-spread ice-wedge degradation is transforming low-centered polygons into high-centered polygons at an alarming rate. Accurate data on spatial distribution of ice-wedge polygons at a pan-Arctic scale are not yet available, despite the availability of sub-meter-scale remote sensing imagery. This is because the necessary spatial detail quickly produces data volumes that hamper both manual and semi-automated mapping approaches across large geographical extents. Accordingly, transforming big imagery into ‘science-ready’ insightful analytics demands novel image-to-assessment pipelines that are fueled by advanced machine learning techniques and high-performance computational resources. In this exploratory study, we tasked a deep-learning driven object instance segmentation method (i.e., the Mask R-CNN) with delineating and classifying ice-wedge polygons in very high spatial resolution aerial orthoimagery. We conducted a systematic experiment to gauge the performances and interoperability of the Mask R-CNN across spatial resolutions (0.15 m to 1 m) and image scene contents (a total of 134 km²) near Nuiqsut, Northern Alaska. The trained Mask R-CNN reported mean average precisions of 0.70 and 0.60 at thresholds of 0.50 and 0.75, respectively. Manual validations showed that approximately 95% of individual ice-wedge polygons were correctly delineated and classified, with an overall classification accuracy of 79%. Our findings show that the Mask R-CNN is a robust method to automatically identify ice-wedge polygons from fine-resolution optical imagery. Overall, this automated imagery-enabled intense mapping approach can provide a foundational framework that may propel future pan-Arctic studies of permafrost thaw, tundra landscape evolution, and the role of high latitudes in the global climate system. Full article

Ice-wedge polygons are widespread and conspicuous surficial expressions of ground-ice in permafrost landscapes. Thawing of ice wedges triggers differential ground subsidence, local ponding, and persistent changes to vegetation and hydrologic connectivity across the landscape. Here we characterize spatio-temporal patterns of ice-wedge degradation since circa 1950 across environmental gradients on Alaska’s North Slope. We used a spectral thresholding approach validated by field observations to map flooded thaw pits in high-resolution images from circa 1950, 1982, and 2012 for 11 study areas (1577–4460 ha). The total area of flooded pits increased since 1950 at 8 of 11 study areas (median change +3.6 ha; 130.3%). There were strong regional differences in the timing and extent of degradation; flooded pits were already extensive by 1950 on the Chukchi coastal plain (alluvial-marine deposits) and subsequent changes there indicate pit stabilization. Degradation began more recently on the central Beaufort coastal plain (eolian sand) and Arctic foothills (yedoma). Our results indicate that ice-wedge degradation in northern Alaska cannot be explained by late-20th century warmth alone. Likely mechanisms for asynchronous onset include landscape-scale differences in surficial materials and ground-ice content, regional climate gradients from west (maritime) to east (continental), and regional differences in the timing and magnitude of extreme warm summers after the Little Ice Age. Full article