Remote sensing of biogeophysical variables at the Cape Bounty Arctic Watershed Observatory, Melville Island, Nunavut, Canada

Abstract

The Cape Bounty Arctic Watershed Observatory (CBAWO), Melville Island, Nunavut (74°55′N, 109°34′W) was established in 2003 to examine Arctic ecosystem processes that would be impacted by climate warming and permafrost degradation. This paper provides a synthesis of how remote sensing has contributed to biogeophysical modelling and monitoring at the CBAWO from 2003 to 2023. Given the location and isolated nature of the CBAWO in the Canadian High Arctic, remote sensing data and derivatives have been instrumental for studies examining ecosystem structure and function at local and landscape scales. In combination with field measurements, remote sensing data facilitated mapping and modelling of vegetation types, % vegetation cover and aboveground phytomass, soil moisture, carbon exchange rates, and permafrost degradation and disturbance. It has been demonstrated that even in an environment with limited vegetation cover and phytomass, spectral vegetation indices (e.g., the normalized difference vegetation index) are able to model various biogeophysical variables. These applications are feasible for research sites such as the CBAWO using high spatial resolution remote sensing data across the visible, infrared, and microwave regions of the electromagnetic spectrum. Furthermore, as the satellite record continues to expand, we will gain a greater understanding of the impacts arising from the expected continued warming at northern latitudes. Although the logistics for research in the Arctic remain challenging, today's technologies (e.g., high spatial resolution satellite remote sensing, automated in situ sensors and data loggers, and wireless communication systems) can support a host of scientific endeavours in the Arctic (and other remote sites) through modelling and monitoring of biogeophysical variables and Earth surface processes with limited but critical field campaigns. The research synthesized here for the CBAWO highlights the essential role of remote sensing of terrestrial ecosystems in the Canadian Arctic.

Introduction

Arctic terrestrial (i.e., non-glacial) ecosystems cover approximately 5 million km² globally, with almost half of this area falling within Canada's borders (i.e., 2.4 million km²) (Walker et al. 2005). These high-latitude environments are important for climate change research given they are impacted by Arctic amplification (i.e., the enhanced increase of near-surface air temperatures over the Arctic relative to lower latitudes) (Ono et al. 2022). Enhanced temperatures in the Arctic are largely attributed to cryospheric changes, in particular the reduction of snow cover and sea ice extent. Any positive feedback mechanisms associated with reduced albedo across the Arctic will further enhance near-surface temperatures (for a detailed examination of Arctic amplification, see Previdi et al. 2021). As a result, the Arctic is one of Earth's most susceptible regions to climate change (Shu et al. 2022), yet the response of these critically important systems to warming remains inadequately understood (Diepstraten et al. 2018; Garcia Criado et al. 2023). It has been noted that while the mean annual Arctic Amplification Index varied between 2 and 3 times the global average from 1970 to 2000, it continues to increase (Screen and Simmonds 2010; AMAP 2019). Chylek et al. (2022) have reported that the Arctic Amplification Index has exceeded four times the global average in the last two decades (i.e., 2000–2020). This finding is supported by others examining temperature anomalies across high latitudes (e.g., Isaksen et al. 2022; Rantanen et al. 2022). Although variable across the Arctic, Box et al. (2019) reported that Arctic annual air temperatures increased by 2.7 °C from 1971 to 2017 and by 3.1 °C for the cold season (i.e., October to May). Temperature increases in the Arctic are outpacing the global mean (IPCC 2021) and will differentially impact the cryosphere overall, with severe ramifications for the Earth system. For example, recent global climate models project that increased global temperatures will be coupled with increased precipitation due to enhanced evaporation and the poleward transport of moisture, transitioning the Arctic from a snow- to a rain-dominated environment over time (Bintanja 2018; McCrystall et al. 2021).

Given a rapidly changing climate in the Arctic, there is an urgency to expand our knowledge of the processes and impacts of these changes on Arctic ecosystems (Lamoureux and Lafrenière 2017). Researchers at the Cape Bounty Arctic Watershed Observatory (CBAWO) have adopted a watershed approach to understanding the impacts of climate change on Arctic ecosystems (i.e., quantifying inputs and outputs to examine watershed-ecosystem responses to environmental change). This approach allows for the integration of terrestrial and aquatic process studies into a holistic framework to examine structure, function, and change within a defined watershed/catchment (Likens 2001). Permafrost is highly susceptible to increased temperatures and moisture inputs, giving rise to unstable landscapes with increased disturbances impacting watershed function (i.e., active layer detachments (ALDs), retrogressive thaw slumps (RTSs), and thermokarst lakes and ponds). These thermokarst features vary in size and frequency across the landscape based on underlying terrain conditions (e.g., slope, aspect, ice content) and are often the first indicators of a changing environment. Further, this warming trend has widespread and diverse impacts on Arctic vegetation, which is predicted to undergo continued shifts in the future (Chapin et al. 2005; Myers-Smith et al. 2011; Post et al. 2019; Berner et al. 2020; Heijmans et al. 2022; Ogden et al. 2023). Plant growth is predicted to increase and affect plant phenology, abundance, biodiversity, and reproductive success, thereby changing community boundaries, composition, and overall ecosystem processes (i.e., carbon exchange, productivity, and energy balance) (Elmendorf et al. 2012; Bjorkman et al. 2017; Gaspard et al. 2023). Unfortunately, we still do not comprehensively understand the interactions between positive and negative feedback associated with warming or their impacts on Arctic terrain and vegetation due to limited and sporadic observations across the Arctic (Goosse et al. 2018; Ogden et al. 2023). Hence, long-term monitoring of Arctic ecosystem patterns, structures, processes, and functions is critical to determining their responses to this warming phenomenon.

Remote sensing data collected across the electromagnetic spectrum and at multiple spatial resolutions using passive and active sensors provide important metrics for examining biogeophysical variables as well as permafrost degradation and disturbance across local, landscape, and regional scales. In turn, these remote sensing data require field measures to train and validate variable models. Over the last two decades, and to support research examining the physical and ecological processes governing ecosystem structure and function, we have conducted remote sensing research across a latitudinal gradient in the Canadian Arctic: Sabine Peninsula, Melville Island, NU, 76° 30′N, 108°50′W (e.g., Liu et al. 2017; Rudy et al. 2017); the CBAWO, Melville Island, NU, 74°55′N, 109°34′W (e.g., Rudy et al. 2013; Freemantle et al. 2020); Lord Lindsay River Watershed, Boothia Peninsula, NU, 70°11′N, 93°44′W (e.g., Laidler et al. 2009; Atkinson and Treitz 2012); Apex River Watershed, Iqaluit, NU, 63°48′N, 68°31′W (e.g., Edwards and Treitz 2017; Liu and Treitz 2018); and the forest-tundra ecotone, Northwest Territories, ∼65°N, 113°W (Bonney et al. 2018). Collectively, these sites represent a gradient of temperature and vegetation ranging from Phytogeographic/Bioclimatic Subzones A (Cushion Forb), B (Prostrate Dwarf Shrub), C (Hemi-prostrate Dwarf Shrub), and D (Erect Dwarf Shrub) (Walker et al. 2005) to the forest-tundra ecotone traversing the 1:1000 and 1000:1 tree:tundra ratio lines (Timoney et al. 1992; Ecosystem Classification Group 2008, 2012). These studies have provided test cases and allowed for comparative analyses of High Arctic conditions at the CBAWO. While the goal here is to synthesize how remote sensing has supported two decades of environmental monitoring at the CBAWO, this synthesis is positioned within this broader latitudinal context. Satellite-based optical and microwave (i.e., SAR) remote sensing data have been collected at the CBAWO and analyzed to (i) generate vegetation type classifications; (ii) model biogeophysical variables (i.e., % vegetation cover (PVC), aboveground phytomass (AGP), soil moisture, ecosystem carbon dioxide exchange rates (i.e., gross ecosystem exchange (GEE), ecosystem respiration (ER), net ecosystem exchange (NEE)); and (iii) map and model permafrost disturbance and degradation. These analyses have been conducted at varying timescales, from a single point-in-time to time-series analyses over multiple years.

The Cape Bounty Arctic Watershed Observatory

The CBAWO is located on the south-central coast of Melville Island, Nunavut (74°55′ N, 109°34′ W) and is approximately 400 km from the nearest community of Resolute, Nunavut (Fig. 1). The research site was established in 2003 to examine hydrological processes for a High Arctic watershed that would be impacted by climate warming and permafrost change, thereby affecting seasonality and the magnitude of associated fluxes (Lamoureux and Lafrenière 2017). Over the past 20 years, remote sensing data and field measures have been collected during wide-ranging climate conditions, including cold to warm melt-season conditions and the warmest decade on record (2007–2016) (Lamoureux and Lafrenière 2017). In short, the CBAWO has experienced significant environmental change, transitioning from a nival (snowmelt)-dominated hydrological regime to an environment increasingly influenced by rainfall runoff and baseflow during the summer season (Lamoureux and Lafrenière 2017; Beel et al. 2021). Research at the CBAWO has since expanded and encompasses studies on hydrological processes (e.g., Favaro and Lamoureux 2014; Bolduc et al. 2018); limnology (e.g., Stewart and Lamoureux 2011; Dugan et al. 2012; Normandeau et al. 2016), paleoclimate (e.g., Cuven et al. 2010, 2011), aquatic ecology (e.g., Stewart and Lamoureux 2012; Roberts et al. 2017), fluvial fluxes of dissolved and particulate materials (e.g., Cockburn and Lamoureux 2008; Fouché et al. 2017; Beel et al. 2020), greenhouse gas exchange (e.g., Wagner et al. 2019; Wright et al. 2021; Hung et al. 2023), permafrost dynamics (e.g., Favaro and Lamoureux 2015; Fouché et al. 2020)), and others.

[Image omitted: See PDF]

The CBAWO consists of paired watersheds that flow into separate “West” and “East” Lakes (unofficial names), which subsequently drain into Viscount Melville Sound (Fig. 1). The “West” and “East” watersheds are approximately 9.5 km² and 11.4 km² in size, respectively, and are underlain by continuous permafrost with an active layer ranging from 0.5 to 1 m depth by late summer (Rudy et al. 2013). The size and proximity of these two watersheds provide a “control” scenario for scientific studies examining watershed processes. Hodgson et al. (1984) describe the bedrock for this region of Melville Island as Paleozoic near-shore marine sandstone and siltstone, while England et al. (2009) indicate the region is underlain by pre-late Wisconsinan glacial till. Glacial and early Holocene marine sediments, which are generally high in silt and clay content, low in organic matter, and poorly drained, cover the area. Finally, the landscape can be described as possessing incised low-elevation plateaus and gentle hills with large areas of exposed and fractured bedrock (i.e., felsenmeer) (Hodgson et al. 1984; England et al. 2009).

The CBAWO is characterized by a polar semi-desert to desert climate with a mean annual air temperature of −14.8 °C (±1.3 °C) and low annual precipitation (<150 mm) (Beel et al. 2021). As a result, the growing season is short (i.e., mid-June to mid-August). It should be noted that increased late summer rainfall at the CBAWO over the past two decades has impacted the hydrological regime as well as the terrestrial-aquatic connectivity for fluxes of dissolved and particulate matter (Beel et al. 2021). Vegetation growth and distribution are strongly influenced by moisture gradients, giving rise to a mosaic of vegetation types ranging from polar deserts to wet sedges. Climate in the watershed is strongly influenced by its proximity to sea ice, with mean summer (June–July–August) air temperatures ranging from 1 to 5 °C between 2003 and 2018 (Beel et al. 2018).

High Arctic vegetation patterns and processes are unique and are not considered analogous to temperate vegetation (Griggs 1934; Beschel 1970). Ritchie (1993, p. 102) expands on this sentiment by stating that the concept of vegetation zonation is of limited use in the High Arctic, given that vegetation composition and structure appear to be controlled by the “maintenance of moist to wet soil surfaces throughout the summer”. Soil texture and the physical and chemical properties of the parent material contribute further to the mosaic of plant communities (Bliss et al. 1984; Edlund and Alt 1989). These authors describe the vegetation as polar desert and polar semi-desert and attribute their distribution to microtopography, rock types, and moisture availability. The primary vegetation that covers the CBAWO is a graminoid, prostrate dwarf shrub, forb tundra (one of 15 physiognomic units based on plant growth forms in the circumpolar Arctic) (CAVM Team 2003; Walker et al. 2005, 2018, 2023). However, it is known from field studies at the CBAWO (and supported by surveys on Melville Island by Bliss and Svoboda (1984) and Bliss et al. (1984)) that the surface cover is quite heterogeneous but tightly coupled to a moisture gradient that is largely a function of topography (and proximity to permanent snowbeds) (Atkinson and Treitz 2012). The most prominent species observed at the CBAWO include: (i) prostate dwarf shrubs (e.g., Salix arctica (Arctic willow), Cassiope tetragona (Arctic white heather), Dryas integrifolia (Arctic mountain avens)); (ii) herbaceous tundra (e.g., Papaver radicatum (Arctic poppy), Saxifraga oppositifolia (Purple saxifrage)); (iii) grasses and sedges (e.g., Poa sp., Eriophorum sp. (Cottongrass)); (iv) lichens (e.g., Umbilicaria sp. (Rock Tripe), Rizocarpon geographicum (Map lichen), Thamnolia subuliformis (Worm lichen)); (v) mosses (e.g., Bryum sp.); and (vi) cyanobacteria (e.g., Nostoc commune) (Brassard 1967; Edlund 1983; LaFarge-England 1989; Atkinson and Treitz 2012).

Vegetation type classification

Generating land-cover classifications for the Arctic is necessary to understand current ecosystem conditions. More specifically, researchers examining hydrological processes, surface water chemistry, permafrost degradation, greenhouse gas exchange, etc., require baseline land-cover information to develop robust sampling schemes. And given the anticipated impacts of enhanced warming on sensitive Arctic landscapes (e.g., vegetation type change and permafrost degradation and disturbance), land-cover classifications offer an effective means of monitoring these changes over time. However, as a result of challenging and costly logistics for conducting research in the Arctic (Mallory et al. 2018), mapping of Arctic landscapes tends to be infrequent, restricted to local-scale studies, and/or of low spatial resolution. Yet, land-cover information remains critical to interdisciplinary studies of Arctic ecosystem structure and function (Bartsch et al. 2016).

There are various motivations and methods for classifying vegetation cover (Desjardin et al. 2023). Ecologists have developed precise ecosystem classifications at various spatial scales that incorporate detailed information based on vegetation taxonomy as well as soils, which are aligned closely with soil moisture and nutrient regimes. The Canadian Arctic-Subarctic Biogeoclimatic Ecosystem Classification (CASBEC) is a detailed ecosystem classification incorporating taxonomic composition and ecological community structure (McLennan et al. 2018). This classification scheme possesses a robust methodology detailing species and environmental conditions that can be linked with remote sensing data (e.g., Ponomarenko et al. 2019). Although critical to understanding ecosystem processes at local scales, the field data required for these classifications tend to be intense, time-consuming to collect, and not always necessary for large-area vegetation classification. For example, a simple vegetation classification may be sufficient to develop sampling schemes to examine a range of science objectives. Finally, given the expanse and the remote nature of the Arctic, there is a lack of baseline data with respect to weather, land cover, and terrain. Weather stations tend to be sparse, and land cover and terrain information is poorly documented and generally unavailable in a high-resolution, spatially explicit form. Hence, there is a need for a classification methodology that is relatively simple to generate in a short period of time while at the same time being ecologically relevant (i.e., related to environmental variables/gradients such as the moisture regime).

Satellite remote sensing data provide spatially explicit multispectral reflectance data that can be analyzed to generate land-cover classifications (including vegetation types) at multiple scales, balancing spatial coverage and taxonomic precision. The Circumpolar Arctic Vegetation Map (CAVM) (CAVM Team 2003) represents the primary mapping initiative that has provided synoptic, spatially explicit vegetation information at the pan-Arctic scale (1:7 500 000), recently upgraded to a 1 km spatial resolution product (Raynolds et al. 2019; CAVM Team 2023). Although helpful for studying large areas, the CAVM is limited when attempting to study local processes related to terrain and vegetation. For instance, the vegetation at the CBAWO is represented by a single physiognomic unit (i.e., graminoid, prostrate dwarf shrub, forb tundra). Given their spatial coverage, frequency of collection, and availability, intermediate spatial resolution satellite data (e.g., Landsat, Sentinel, and SPOT) have been analyzed for characterizing land cover at high latitudes (Macander et al. 2017; Langford et al. 2019) and will be necessary for upscaling ecosystem conditions and processes to regional scales (Rudd et al. 2021). However, these data present their own challenges as the spatial and spectral resolution remains limited for in situ studies; hence, the discrimination of vegetation types at high spatial resolution remains challenging.

Over the last two decades, high spatial and spectral resolution satellite remote sensing data have become more readily available (e.g., Quickbird, IKONOS, GeoEye, WorldView, and Planet) for conducting more precise assessments of land cover and vegetation (Virtanen and Ek 2014), with the potential to quantify vegetation types and biogeophysical variables that control ecosystem processes (e.g., PVC, AGP, GEE, ER, and NEE) at local spatial scales and over short time scales (Reichle et al. 2018). These sensors (and platforms) continue to evolve to provide enhanced spatial and spectral information, along with data for atmospheric correction. And now, with the proliferation of remotely piloted aircraft systems (RPASs), local- and landscape-scale imaging is becoming more common, providing sub-decimeter spatial resolution for multiple spectral channels (Fraser et al. 2016; Stanski et al. 2021; Thomson et al. 2021; Sluijs et al. 2023; Wolff et al. 2023). Since 2003, we have been routinely collecting high spatial resolution satellite data during the summer season at the CBAWO (i.e., IKONOS, GeoEye, and WorldView). These data have made it feasible to collect precise land-cover information critical for examining local- and landscape-level ecosystem processes. They also allow for capturing more discrete thermokarst activity across the landscape (e.g., ALDs and RTSs) and thermokarst ponds (Rudy et al. 2013; Jones et al. 2023). For a comprehensive survey of remote sensing platforms and sensors, the reader is referred to Toth and Jóźków (2016).

Various land-cover classifications have been generated for the CBAWO to support field studies examining vegetation and earth-surface processes. Initially, Gregory (2011) developed a general five-class land-cover classification (i.e., polar semi-desert, mesic tundra, wet sedge meadow, bare ground, and water) by applying a rules-based methodology to separate supervised maximum likelihood classifications and normalized difference vegetation indices (NDVI) of IKONOS multispectral (4 m spatial resolution) data collected on 4 July and 2 August 2008 (83% overall accuracy; Kappa coefficient = 0.79). In an effort to integrate ecosystem characteristics and spectral reflectance data, Atkinson and Treitz (2012) developed a methodology to combine field biogeophysical measurements and IKONOS spectral reflectance data collected on 22 July 2004 to generate an ecological classification. They used an ordination analysis that applied species abundance data to generate spectral classes that were submitted to a maximum likelihood classification. Species and cover abundance (i.e., % cover of biotic species (genus in the case of mosses) and abiotic components such as rock and soil) were estimated using the Braun-Blanquet scale (Braun-Blanquet 1932). This analysis revealed the relationship between ecological data (i.e., species abundance) and environmental variables (i.e., soil moisture, exposed soil, and rock) that were critical to class distribution. The resulting classification (Dry Papaver radicatum and Till Tundra, Mesic Nostoc commune Tundra, Wet Sedge Tundra, Felsenmeer, Water, and Snow) exhibited an overall accuracy of 79% (Kappa Coefficient = 0.69).

Most recently, and as satellite remote sensing technology has evolved, a high spatial resolution environmental land-cover classification (i.e., 2 m spatial resolution) was developed for the CBAWO using 8-band WorldView-2 visible–near-infrared (VNIR) top-of-atmosphere (TOA) reflectance data collected in 2016 (Figs. 1–3) and a topographic wetness index (TWI) (Hung and Treitz 2020). TWI, a proxy for soil moisture (Wilson and Gallant 2000), is a terrain variable derived from slope, flow direction, and flow accumulation metrics modelled from a digital elevation model (DEM). Here, the TWI data were derived from a 1 m resolution DEM generated from a WorldView-2 panchromatic stereo-pair (0.5 m spatial resolution) (Collingwood 2014) (Fig. 4). This further demonstrates the utility of high spatial resolution remote sensing satellite data for generating valuable terrain surfaces (e.g., elevation, slope, aspect, and TWI) for watershed research. [Note: High resolution digital elevation data are now readily available for the circumpolar Arctic due to the development and continued support of the ArcticDEM https://www.pgc.umn.edu/data/arcticdem/ (Porter et al. 2018; Karlson et al. 2021).] Although a specific classification algorithm (i.e., support vector machine (SVM)) was employed to generate this classification, there are many suitable classification algorithms available to perform this task. However, machine learning algorithms (e.g., SVM, random forests (RFs), and artificial neural networks (ANNs)) are good candidates for classification given that they are robust when analyzing diverse datasets of different measurement scales (e.g., interval, ratio) or distribution types (parametric versus non-parametric). This CBAWO classification possesses a precise classification scheme with a high overall classification accuracy (i.e., 90.7%; Kappa coefficient = 0.89) (Hung and Treitz 2020). Classification products have proven critical to the research community at the CBAWO by connecting land cover with hydrological outputs (e.g., Lewis et al. 2012; Holloway et al. 2016; Fouche et al. 2017) and greenhouse gas fluxes (e.g., Wagner et al. 2019; Atkinson et al. 2020; Braybrook et al. 2021; Wright et al. 2021). Further, an accurate and precise environmental classification will better inform an integrated watershed approach for upscaling of site-level biogeophysical variables.

[Image omitted: See PDF][Image omitted: See PDF][Image omitted: See PDF]

Biogeophysical variables

In the High Arctic, air and soil temperature, soil moisture, nutrient availability, active layer depth, topography, micro-topography, and soil type serve as controls affecting vegetation growth (Evans et al. 1989; van Wijk and Williams 2005; Wookey et al. 2009; Walker et al. 2011, 2016; Hobbie et al. 2017; McLennan et al. 2018; Mikola et al. 2018; Beamish et al. 2020). Here we outline how various remote sensing data and methods have contributed to quantifying key biogeophysical variables at the CBAWO.

Percent vegetation cover

PVC, or fraction of green vegetation, an important indicator of ecosystem health and productivity, is defined by Purevdorj et al. (1998) as the percentage of the ground surface covered by green vegetation. A key derivative of spectral data for characterizing PVC in Arctic environments is the NDVI (e.g., Laidler et al. 2009; Epstein et al. 2012; Raynolds et al. 2012, 2013; Liu and Treitz 2016). This results from an important relationship between near-infrared (NIR) wavelengths (700–1200 nm, which are highly reflective from leaf spongy parenchyma mesophyll cells within leaves) and red (R) wavelengths (600–700 nm, which are highly absorbed by chlorophylls a and b present within the palisade parenchyma mesophyll cells) (eq. 1) (Gaussman et al. 1969; Rouse et al. 1973; Tucker 1979; Jensen 2016):

[Formula omitted: See PDF]

where ρ is % reflectance. Many other vegetation spectral indices and metrics have been developed for examining specific absorption features related to plant foliar traits, e.g., chlorophyll, nitrogen, and lignin (Treitz et al. 2010; Liu et al. 2017; Wang et al. 2020), but many are modifications of the core spectral relationship between chlorophyll absorption and leaf reflectance within the visible and near-infrared spectral regions, respectively.

Atkinson and Treitz (2013) determined that NDVI data derived from IKONOS data collected on 22 July 2004 were closely correlated to PVC as measured in the field using the Braun-Blanquet scale (Braun-Blanquet 1932) at the CBAWO and on Boothia Peninsula (i.e., PVC = 0.024 (NDVI) – 0.073; r² = 0.88, p < 0.001) (Fig. 5). Although the observed relationship between NDVI and dry AGP was strong on the Boothia Peninsula (i.e., r²= 0.76), it was not as pronounced at the CBAWO (i.e., r² = 0.55). This may have been a result of the spatial distribution of phytomass in the polar semi-desert class, with large areas of bare soil and more concentrated vegetation at the edges of the polygon wedges. Finally, the authors demonstrated a strong relationship between NDVI and peak growing season soil moisture (i.e., r² = 0.72), again illustrating the dependence of vegetation distribution and abundance on the moisture regime (Bliss and Matveyeva 1992; Nobrega and Grogan 2008). This supports the premise that the most precise vegetation classification derived for the CBAWO (Fig. 2) exhibits a strong surface/soil moisture gradient.

[Image omitted: See PDF]

Liu and Treitz (2016) examined various normalized difference spectral indices derived from WorldView-2 visible and near-infrared wavelengths (i.e., yellow, red, red edge, and near-infrared) and found strong correlations to PVC derived from classified field photographs. The same authors also determined that image-based estimates of field PVC were highly correlated to point-frame estimates of PVC (R² > 0.80), indicating that image techniques can potentially replace time-consuming field measurements to minimize the logistical challenges of extended field data collection at remote locations (Liu and Treitz 2018). This technique was also used successfully by Sellers et al. (2023). It is clear that biophysical variables related to vegetation amount (i.e., PVC, AGP) can be modelled using spectral indices that are typically based on the relationship between chlorophyll concentration and cell structure and reflectance in the visible and near-infrared. Given the spatial heterogeneity of vegetation at this High Arctic site, high spatial resolution remote sensing data are ideal for characterizing biophysical variables related to vegetation structure.

Aboveground phytomass

Vegetation is very limited in the Canadian High Arctic (Raynolds et al. 2006; Walker et al. 2012). Albeit limited, vegetation cover and AGP have significant impacts on the terrestrial carbon balance (Boelman et al. 2003; Shaver et al. 2007) and provide food for Arctic ungulates such as Peary Caribou (Rangifer tarandus pearyi) and muskoxen (Ovibos moschatus) (Parker and Ross 1975; Abuelgasim and Leblanc 2011; Maher et al. 2012). Collingwood et al. (2014a) developed a novel approach using ANNs to model AGP (g/m²). First, ANNs were created that defined the relationship between high-resolution multi-incidence angle RADARSAT-2 SAR data and GeoEye-1 soil-adjusted vegetation index (SAVI) data. Although the SAVI model results for individual ecological classes were modest (i.e., semi-desert (r²= 0.43), mesic heath (r²= 0.43), wet sedge (r²= 0.30), and felsenmeer (r²= 0.59)), the r² improved to 0.60 when the model outputs were combined to analyze the relationship between the model output and SAVI as a whole (with a low normalized root mean square error (RMSE) of 8%) (Fig. 6). This SAR relationship to SAVI can be exploited given the persistent cloud cover conditions typical of High Arctic environments. More significantly, the AGP model exhibited a very strong relationship for SAR-modelled and field-measured AGP (i.e., r²= 0.87) that exceeded the relationship between optically derived SAVI values and field-measured AGP (i.e., r²= 0.79). Total AGP was clipped within each 50 cm × 50 cm quadrat used to characterize PVC and AGP for the three primary vegetation types (polar semi-desert, mesic heath, and wet sedge tundra). Field AGP samples were weighed in the field, air dried, and then shipped to the lab for oven drying (Atkinson and Treitz 2013; Collingwood et al. 2014a). These results are unique in that they demonstrate the utility of SAR data for modelling the relatively low levels of AGP at the CBAWO.

[Image omitted: See PDF]

Soil moisture

Soil moisture is an important biogeophysical variable impacting hydrological patterns and processes (Grayson and Blöschl 2001) and is a key variable for understanding Arctic energy fluxes (Crosson et al. 2002). It has also been demonstrated that soil moisture influences patterns of CO₂ exchange rates at high latitudes (Illeris et al. 2003; Oberbauer et al. 2007; Dagg and Lafleur 2011; Reichstein et al. 2013). It is anticipated that climate warming will impact soil moisture, but this impact is expected to be highly spatially variable. This variability is important to capture at local and landscape scales as it will differentially impact vegetation development and distribution, active layer depth development, and wetlands (Woo and Xia 1995; Hinkel et al. 2001; Natali et al. 2012). In short, soil moisture and associated hydrologic variables are important controls on Arctic geomorphology, permafrost, and ecosystem dynamics, and therefore an important variable to model in the Arctic.

The Canadian Government is keenly aware of the challenges of obtaining optical data at high latitudes due to solar patterns and extensive cloud cover, and therefore has invested in synthetic aperture radar (SAR) satellite technologies (i.e., Radarsat I and II and Radarsat Constellation Mission) to monitor these high latitude environments. As a result, we have explored the potential for SAR data to characterize soil moisture at the CBAWO. Initially, Radarsat I data (i.e., C-HH in standard beam mode) were examined to determine sensitivity to soil (and surface) moisture (Wall et al. 2010). Examination of two dates of SAR data (8 July and 1 August 2004) identified a decrease in mean image radar backscatter (sigma nought (σ°)) corresponding to a decrease in soil moisture (i.e., 8.6%) in the top 5 cm of the soil between the two dates. The authors noted that separating the combined effects of vegetation (and organic material for wet sedge), surface roughness, topography, and moisture content on radar backscatter was difficult. Further, one of the most significant findings of this study was the reduced backscatter response from areas exhibiting moisture saturation (specifically near streams and areas with shallow active layers where moisture accumulated). This was attributed to an enhanced specular response from saturated soils that minimized the dielectric effect of water content (Wall et al. 2010). This phenomenon would have effectively dampened the contrast in soil moisture content between the two dates.

Given these results, it was determined that modelling surface roughness (Zs), a time-invariant variable, would be necessary to derive % volumetric soil moisture (Ɵv) adequately. Surface roughness values were measured in the field using a 0.92 m pin meter along 5.5 m transects in two orthogonal directions relative to the SAR satellite look direction (i.e., parallel and perpendicular) (Collingwood et al. 2014b). Pin meter measurements were transformed into root mean square height, which is a measure (in cm) of the vertical variation of roughness and correlation length, which provides a measure of the horizontal variation in roughness. Multiple incidence angle data and fully polarimetric data from RADARSAT-2 were combined with long- and short-profile in situ surface roughness measurements to generate a surface roughness model (Collingwood et al. 2014b). Subsequently, surface roughness (as a static variable) was combined with time-dependent variables, i.e., HH polarized backscatter (σ°) and local incidence angle (Ɵ_l) specific to each image acquisition, to model % volumetric soil moisture (Ɵ_v) derived from time-domain reflectometry sampled across the CBAWO (Collingwood et al. 2018). The results of applying the model to time-series images illustrated the sensitivity to soil moisture changes resulting from rainfall events (Fig. 7). Here, it is demonstrated that Ɵ_v responded to average soil moisture change at the plot level (0.27–0.36 m³/m³) as a result of rainfall events from 12 to 15 July (2.8 mm). This capacity allows for spatially explicit soil moisture modelling over time at high spatial resolution. The soil moisture model developed at the CBAWO was robust (absolute soil moisture errors of 15% (r²= 0.46)) with different input beam modes across a range of physical conditions and moisture regimes. It was then validated for the Sabine Peninsula, Melville Island, exhibiting strong results (i.e., an absolute soil moisture error of 12% (r²= 0.26)), indicating that modelling efforts of this nature may be transferrable, at least regionally, in the High Arctic. The authors also determined that the model was scalable across various size image objects, thereby giving potential for upscaling across larger areas (Collingwood et al. 2018).

[Image omitted: See PDF][Image omitted: See PDF]

Ecosystem carbon exchange

Much of the interest in northern environments stems from the impacts of Arctic amplification and possible feedback mechanisms that can either enhance or reduce warming. One significant potential feedback is the relationship between increased temperature, increased respiration, and/or increased ecosystem productivity—hence the need to understand, model, and monitor CO₂ exchange rates across spatial and temporal scales (Virkkala et al. 2018, 2023). For instance, elevated air temperatures are likely to increase growing degree days and extend growing season lengths, resulting in increased productivity (i.e., photosynthesis, AGP, and GEE) of vegetation in the Arctic (Braybrook et al. 2021). However, coincident with increased GEE is enhanced ER (i.e., autotrophic and heterotrophic respiration from plants and soils, respectively). Further, and as a function of vegetation, soil, and environmental variables (e.g., vegetation type, soil organic matter, soil moisture, nutrient availability, macro- and micro-topography, temperature, and thaw depth), it has been demonstrated that GEE and ER vary in relation to vegetation type (Nobrega and Grogan 2008; Dagg and Lafleur 2011; Emmerton et al. 2016; Atkinson et al. 2020). As a result, to understand carbon exchange rates across the High Arctic, it is important to examine the variability in NEE among different vegetation types. Remote sensing can help quantify and map carbon exchange rates by mapping vegetation types or quantifying key biophysical properties (e.g., foliar properties through the NDVI) that influence carbon exchange rates. This will assist in estimating carbon budgets at local to landscape scales.

Arctic ecosystems store tremendous amounts of organic material (i.e., carbon) in Arctic permafrost soils, which will be susceptible to aerobic or anaerobic respiration upon thawing, thereby releasing CO₂ or methane (CH₄), respectively, potentially creating positive permafrost carbon feedback (Miner et al. 2022). Given the heterogeneity of Arctic terrestrial environments, it is important to quantify the carbon source and sink strength (i.e., NEE) across a variety of high-latitude ecosystem types. These measurements are typically conducted using static or automated chambers (e.g., Acosta et al. 2013; Pavelka et al. 2018) or using eddy covariance (e.g., Lafleur and Humphreys 2018; Richardson et al. 2019). However, eddy covariance measures may include multiple vegetation types in the footprint of the tower, making it difficult to attribute carbon exchange rates to specific vegetation types and complicating the extrapolation of carbon exchange spatially. Therefore, if remotely sensed measures of vegetation type/quality can be coupled with chamber measurements, it remains possible to extrapolate these exchange rates over space, at least for short intervals during peak summer growth for contrasting vegetation types.

At the CBAWO, we have found good relationships between NDVI and the average seasonal CO₂ estimates of GEE, ER, and NEE (Atkinson et al. 2020; Braybrook et al. 2021). Braybrook et al. (2021) synthesized 5 years of eddy covariance results to partition NEE into its component fluxes of GEE and ER. Cumulative growing season carbon exchange was modelled with single growing season NDVI data (IKONOS and Worldview-2). NDVI showed significant correlations with NEE (R² = 0.85) and GEE (R² = 0.80), highlighting the capability to extract landscape-scale seasonal carbon balances using a single-satellite remote sensing product. Atkinson et al. (2020) utilized static chamber data from two sites (Lord Lindsay River Watershed and the CBAWO) to model carbon exchange across a range of vegetation types (but independent of vegetation types) based solely on the spectral reflectance characteristics of vegetation across the watershed (i.e., NDVI). While NDVI is generally considered a function of leaf area or phytomass, at the CBAWO, it appears to be a strong surrogate for summer season carbon exchange rates. This allowed us to derive spatial models of carbon exchange rates at high spatial resolution, thereby demonstrating that a single satellite image captured during peak phytomass can be used to quantify CO₂ exchange across whole watersheds (Fig. 8). Finally, given that CO₂ exchange processes could be predicted for vegetation arrayed across a strong moisture gradient (i.e., dry, mesic, and wet tundra), NDVI exhibits a strong functional convergence for vegetation types. Interestingly, NDVI variability within a vegetation type (e.g., wet sedge) did not correlate with high-frequency (i.e., 30 min) CO₂ exchange processes (Wright et al. 2021), suggesting an important temporal scale at which vegetation properties influence carbon exchange rates. Unfortunately, we do not have CO₂ exchange rate data for the CBAWO for spring, fall (e.g., Nobrega and Grogan 2008; Commane et al. 2017), or winter (Celis et al. 2017; Natali et al. 2019; Watts et al. 2021). Extending environmental variables and CO₂ exchange rate measurements into these seasons is still required to determine the impacts of a changing climate on the annual NEE.

Integrating biogeophysical variables for landscape classification

While the biogeophysical variables modelled at the CBAWO were typically developed and examined individually, there are also potentially useful outputs when they are analyzed in combination. For instance, the model results from the surface roughness, vegetation phytomass, and soil moisture models can be used to represent the physical landscape through the use of a red–green–blue colour space (Fig. 9). This is an easy way to visualize and even classify the landscape using physically meaningful characteristics, rather than the simple optical reflectance that is typically used for land-cover classification. Classification could also include ancillary terrain variables derived from a DEM. With such an extensive library of remotely sensed information available for the CBAWO, consolidating and displaying data in this fashion can yield new insights into the interactions and dependencies of these various biogeophysical variables across the landscape. In fact, terrain variables have been incorporated into the recent land-cover classification (i.e., TWI) (Hung and Treitz 2020) and permafrost disturbance susceptibility models (e.g., TWI, slope (degrees), potential incoming solar radiation (MJ/m²) (Fu and Rich 1999), and topographic position index (Jenness 2006)) (Rudy et al. 2016). Although not yet tested, it is anticipated that SAR-derived variables such as surface roughness and soil moisture could help model permafrost disturbance susceptibility.

[Image omitted: See PDF]

Permafrost disturbance and degradation

Approximately 15 million km² of the exposed land area of the northern hemisphere is underlain by permafrost (Chadburn et al. 2017). As noted above, permafrost stores carbon that can be released through aerobic or anaerobic respiration (i.e., methanogenesis) as permafrost thaws, potentially leading to a significant positive feedback (Koven et al. 2011; Hugelius et al. 2014; Schuur et al. 2015; Plaza et al. 2019). Permafrost thaw has given rise to various forms of degradation and disturbance at the CBAWO and includes thermokarst features such as ALDs and RTSs (Fig. 10). These features vary in scale and are occurring throughout the circumpolar Arctic. Remote sensing data have been examined at the CBAWO to distinguish differing characteristics of permafrost disturbance and surface subsidence (Rudy et al. 2013, 2018; Robson et al. 2021).

[Image omitted: See PDF]

The summer of 2007 was one of the warmest years on record up to that point in time (July mean air temperature = 10.3 °C; 6 °C higher than the climate normal (1971–2000) for the nearest climate station at Mould Bay, NU) (Lafrenière and Lamoureux 2013). Coupled with these exceptional temperatures (which generated a thickened active layer), the CBAWO experienced a significant rainfall event (i.e., 10.8 mm) in late July. These conditions triggered a series of ALDs at the CBAWO (Lamoureux and Lafrenière 2009), particularly in the West catchment (Figs. 1 and 10). Given the variable size and shape of these ALDs, we explored the potential of multi-temporal IKONOS 4 m spatial resolution data to identify and map these features (Rudy et al. 2013). To minimize radiometric differences, IKONOS data were collected on 22 July 2004 and 12 July 2010, relatively similar peak vegetation growth anniversary dates. Still, to model the first-order effects of atmospheric scattering, a relative radiometric correction using dark object subtraction (Hadjimitisis et al. 2010; Jensen 2016) was applied to normalize the intensities for the various bands within each image. Again, NDVI was utilized to detect areas that were vegetated pre-disturbance and exhibited exposed soil and glacial till post-disturbance (Fig. 11). NDVI differencing was applied to examine the differences in vegetation between dates using the following equation:

[Formula omitted: See PDF]

[Image omitted: See PDF]

After co-registering IKONOS 2004 and 2010 NDVI data, the 2010 NDVI data were subtracted from the 2004 data to generate an NDVI difference image to detect zones where vegetation cover decreased and/or was removed entirely (Fig. 11). These zones were then identified and separated from undisturbed locations using object-based image analysis and a fixed NDVI threshold. It was determined that larger and/or elongated ALDs were more readily identified where vegetation removal tended to be more complete, at least for the scar/track zone. As a result, size and morphology were both characteristics of ALDs that impacted their detection using IKONOS data (Rudy et al. 2013). It is anticipated that even higher spatial resolution data (e.g., WorldView) would enhance the detection of smaller ALDs.

Based on the high frequency of occurrence of ALDs at the CBAWO in 2007, the potential for identifying areas of the landscape susceptible to these types of events was explored (Rudy et al. 2016). The CBAWO DEM with vertical and horizontal resolutions of 1 m (Fig. 4) was used to derive the following terrain attributes: slope angle; distance to water; topographic position index; potential incoming solar radiation; and TWI (Beven and Kirkby, 1979). These served as proxies for geomorphic and hydrologic processes determined to be significant for disturbance initiation (Rudy et al. 2016, 2017). As a result, a permafrost disturbance susceptibility model was derived for the CBAWO using generalized linear modelling (Fig. 12). The model applied to the CBAWO was developed based on site characteristics (i.e., disturbed and undisturbed sites) at two other locations (i.e., Sabine Peninsula, Melville Island, NU and Fosheim Peninsula, Ellesmere Island, NU). The CBAWO served as a validation site to test the transferability of the models developed with input data from these locations. For the three regions examined, including the CBAWO, it was determined that terrain attributes associated with the triggering of disturbance were similar (i.e., slopes 4°≤×≤15°; elevation below the Holocene marine limit; and having low potential incoming solar radiation) and that models are transferable if they account for the variability of conditions across regions. However, it was observed that there was an overestimation of high to very high susceptible areas at the CBAWO; this was attributed to the bedrock exposures at the CBAWO, which were not present at the other sites. Models based on the range of site conditions giving rise to disturbance are suitable for developing (and transferring) permafrost susceptibility models across regions (Rudy et al. 2017, 2018). A similar modelling approach successfully identified areas that may be sensitive to high pore water potentials, a precursor condition leading to landscape degradation (Holloway et al. 2017). These models were successful in identifying areas susceptible to ALDs and mud ejections, thereby improving our understanding of geomorphic sensitivity to permafrost degradation.

[Image omitted: See PDF]

Repeated freezing and thawing of the active layer coupled with subsurface ice loss can potentially lead to large-magnitude localized surface subsidence, such as ALDs and RTSs (Burn 2000; Lewkowicz 2007; Kokelj and Jorgenson 2013). Further, more widespread or generalized subsidence occurs seasonally due to thaw consolidation in areas of continuous ice-rich permafrost, such as the CBAWO, but is typically of smaller magnitude (Shiklomanov et al. 2013). Short and Fraser (2023) identify high spatial resolution surface displacement maps (i.e., changes in the state of the permafrost are manifested at the ground surface as surface subsidence or uplift as a function of changes in ground ice content in the active layer and uppermost permafrost (Wolfe et al. 2014)) as useful for infrastructure planning and engineering, while regional-scale overviews and change monitoring can be derived from intermediate scale displacement surfaces. To determine the patterns of surface displacement at the CBAWO, intra- and inter-seasonal displacement surfaces were generated using Differential Interferometric SAR (DInSAR) with ultrafine-beam Radarsat-2 data for the summers of 2013, 2015, and 2019 at the CBAWO (Rudy et al. 2018; Robson et al. 2021). Although intra- and inter-seasonal displacement detected over the three seasons/years was generally less than 4 cm, greater displacements were detected and were typically clustered in wet, low-lying areas and on steep slopes close to the coast (Fig. 13). DInSAR also captured the expansion of RTSs at the CBAWO, exhibiting negative surface change in the slump floors (Robson et al. 2021).

[Image omitted: See PDF]

Conclusions

Our remote sensing research at the CBAWO has focused on the estimation of biogeophysical variables using remote sensing methods that are trained and validated through comprehensive field campaigns. More broadly, our Arctic research has examined biogeophysical variables across a latitudinal gradient, corresponding to a climatic and vegetation gradient—a surrogate continuum for anticipated changes that will be observed under a warming climate (Bonney et al. 2018; Atkinson and Treitz 2012, 2013; Liu and Treitz 2016; Edwards and Treitz 2017, 2018; Atkinson et al. 2020). It is understood that spectral vegetation indices are suited to the prediction of biogeophysical variables, at least at coarse spatial resolutions across large bioclimatic zones in the Arctic (e.g., Raynolds et al. 2012). For our research at the CBAWO, the NDVI was critical for modelling biogeophysical variables (e.g., vegetation type, PVC) at local scales in environments with low PVC and AGP (i.e., the High Arctic). Further, this capability holds true given the diverse range of vegetation types present at the CBAWO (i.e., prostate dwarf shrubs, herbaceous tundra, grasses and sedges, lichens, mosses, and cyanobacteria). We have extended these methods to model ecosystem processes (i.e., GEE, ER, and NEE) and then scaled these processes to landscape scales using high spatial resolution NDVI data. Of importance was the relatively small variation in GEE, ER and NEE measured in the field during the three- to five-week peak growing period, thereby allowing NDVI data derived from a single satellite image to extrapolate these processes spatially.

Although it is well known that surface moisture is a key driver of vegetation productivity and other biogeochemical processes, surface moisture is difficult to extract through conventional means in the High Arctic since it is influenced by active layer depth, microtopography, and permanent snowbeds rather than directly by precipitation. We have developed novel applications using SAR data and ANNs to successfully model biogeophysical variables (e.g., AGP, surface roughness, and soil moisture) in a permafrost environment (Wall et al. 2010; Collingwood et al. 2014a, 2014b, 2018). Further, we have incorporated spectral indices, terrain variables (derived from digital elevation data), and DInSAR into our analyses to model areas susceptible to permafrost disturbance (Rudy et al. 2013, 2016, 2017, 2018). Finally, with regards to longer-term monitoring, our research has determined that terrestrial vegetation changes are spatially variable across latitudes and between vegetation types; i.e., we observed modest “greening” at intermediate spatial resolutions (i.e., using Landsat time series data from 1984 to 2015; Edwards and Treitz 2017) and at high spatial resolutions (i.e., using IKONOS/WorldView time series data from 2004 to 2018; Freemantle et al. 2020).

Remote sensing data and analytical methods have proven instrumental at the CBAWO for enhancing our understanding of terrestrial surfaces and ecosystem processes at local to landscape scales. As satellite sensors continue to evolve and with the application of RPAS at the CBAWO, research will continue to examine the variability of biogeophysical variables within vegetation types to characterize greenhouse gas exchange in response to changes in hydrological processes and permafrost dynamics. As noted above, we are able to generate high resolution spatially explicit terrain variables (e.g., surface roughness, soil moisture, TWI, and potential incoming solar radiation) that can be used to develop more comprehensive classifications and models of land cover, vegetation types, soil moisture, and terrain. DEMs derived from repeat RPAS flights will be able to track the development of dynamic landscape features such as RTSs or estimate any surface deformation related to the seasonal freeze–thaw processes. For instance, in 2018, in conjunction with ground observations, imagery collected with a DJI Phantom 3 RPAS was able to characterize diel and synoptic vertical ground displacement on the order of ±0.15 mm (McFadden 2019). These fine-scale variabilities could be indicators of locations susceptible to larger scale slope failures in the future. In addition, remote sensing of plant phenology for specific vegetation types can be examined more precisely over the growing season in combination with carbon exchange rates to better model seasonal exchange rates. If we can link seasonal carbon exchange rates to detailed vegetation types, we can then model NEE for the CBAWO using vegetation types rather than NDVI. This provides another method for modelling NEE and validating other remote sensing models (i.e., based on NDVI or PVC). As the satellite record continues to grow, researchers at the CBAWO are well positioned to examine changes in productivity and responses to permafrost degradation in the context of long-term environmental change. High spatial resolution remote sensing data have allowed for more precise coupling of field measures to spectral reflectance/radar backscatter, thereby allowing researchers to examine the scale dependency of these processes. Further, a comprehensive analysis of high spatial/spectral resolution remote sensing data determined that narrow-band spectral indices exhibited stronger correlation with PVC than broadband vegetation indices due to the finer spectral features sampled (Liu et al. 2017). In short, high spatial and spectral resolution remote sensing data and derivative products are critical to addressing science objectives in High Arctic permafrost environments and are essential for monitoring environmental change into the future. These data/derivatives will provide more precise estimates of vegetation type, biogeophysical variables, and permafrost disturbance and degradation, particularly given future satellite contributions to these efforts (e.g., hyperspectral sensors, intelligent remote sensing satellite systems; Zhang et al. 2022).

Since 2003, we have had the privilege of conducting our research on the Inuit lands of Melville Island, NU, and we are extremely grateful to the Inuit for that opportunity, particularly the people of the Hamlet of Resolute. Continuous, long-standing localized research activity is rare in the High Arctic, particularly outside of established communities, and as a result, we have been able to study ecosystem processes over time using an integrated watershed approach. Contextualizing Earth system processes within this framework has resulted in a holistic appreciation of how water, carbon, and nutrients move through watersheds and how these processes impact ecosystems and permafrost during warming. However, as noted by Lauden et al. (2017), there are only a small number of Arctic research stations dedicated to examining natural science processes, particularly during this period of comprehensive environmental change arising from warming temperatures. This can be attributed largely to the high cost of conducting research (Mallory et al. 2018), and, as a result, operations have been reduced or closed entirely in recent years. However, without the continued support of these research stations to study and model ecosystem processes within a warming environment, we will not be able to advise on suitable policy initiatives to mitigate environmental change. We continue to be hopeful that governments will recognize that the science generated by personnel at these research stations is critical to inform policy decisions by government agencies.

Acknowledgements

The authors want to thank the Polar Continental Shelf Program (PCSP) for their tremendous logistical support over the last two decades. Valuable research support has been provided over the years by the Natural Sciences and Engineering Research Council (NSERC) Discovery Grants and projects supported by the National Centres of Excellence—ArcticNet, the Government of Canada International Polar Year (IPY), and Environment and Climate Change Canada. The Northern Science Training Program (NSTP) has contributed valuable student support for field activities over the years. Student support from the NSERC, the Weston Family Foundation, and Queen's University is appreciated. We want to express our gratitude to many members of the Resolute, Nunavut community, and the Nunavut Research Institute (NRI) for their continued support for the CBAWO. Over the past two decades, there have been many people who have supported our remote sensing research through field data collection, processing, and analyses. Finally, the principal author would like to thank Dr. Heather Reese and the Department of Earth Sciences at the University of Gothenburg, Sweden, for hosting him during his sabbatical in 2022–2023, where this writing project began.

References

Abuelgasim A.A., Leblanc S.G. 2011. Leaf area index mapping in northern Canada. International Journal of Remote Sensing, 32: 5059–5076.

Acosta M., Pavelka M., Montagnani L., Kutsch W., Lindroth A., Juszczak R., Janouš D. 2013. Soil surface CO2 efflux measurements in Norway spruce forests: comparison between four different sites across Europe-from boreal to alpine forest. Geoderma, 192: 295–303.

Arctic Monitoring and Assessment Program (AMAP). 2019. AMAP climate change update 2019: an update to key findings of snow, water, ice, and permafrost in the Arctic (SWIPA) 2017. AMAP, Oslo, Norway.

Atkinson D., Treitz P.M. 2012. Arctic ecological classifications derived from vegetation community and satellite spectral data. Remote Sensing, 4: 3948–3971.

Atkinson D., Treitz P.M. 2013. Modelling biophysical variables across an Arctic latitudinal gradient using high spatial resolution remote sensing data. Arctic, Antarctic and Alpine Research, 45(2): 161–178.

Atkinson D.M., Hung J.K.Y., Gregory F.M., Scott N.A., Treitz P.M. 2020. High spatial resolution remote sensing models for landscape-scale CO2 exchange in the Canadian Arctic. Arctic, Antarctic, and Alpine Research, 52(1): 1–16.

Bartsch A., Hofler A., Kroisleitner C., Trofaier A.M. 2016. Land cover mapping in northern high latitude permafrost regions with satellite data: achievements and remaining challenges. Remote Sensing, 8(12): 979.

Beamish A., Raynolds M.K., Epstein H., Frost G.V., Macander M.J., Bergstedt H., et al. 2020. Recent trends and remaining challenges for optical remote sensing of Arctic tundra vegetation: a review and outlook. Remote Sensing of Environment, 246: 111872.

Beel C.R., Lamoureux S.F., Orwin J.F. 2018. Fluvial response to a period of hydrometeorological change and landscape disturbance in the Canadian High Arctic. Geophysical Research Letters, 45(19): 10446–10455.

Beel C.R., Lamoureux S.F., Orwin J.F., Pope M.A., Lafrenière M.J., Scott N.A. 2020. Differential impact of thermal and physical permafrost disturbances on High Arctic dissolved and particulate fluvial fluxes. Scientific Reports, 10: 11836.

Beel C.R., Heslop J.K., Orwin J.F., Pope M.A., Schevers A.J., Hung J.K.Y., et al. 2021. Emerging dominance of summer rainfall driving High Arctic terrestrial-aquatic connectivity. Nature Communications, 12: 1448.

Berner L.T., Massey R., Jantz P., Forbes B.C., Macias-Fauria M., Myers-Smith I., et al. 2020. Summer warming explains widespread but not uniform greening in the Arctic tundra biome. Nature Communications, 11: 4621.

Beschel R.E. 1970. The diversity of tundra vegetation. In Productivity and conservation in Northern circumpolar lands. IUCN New Series No. 16. International Union for Conservation of Nature and Natural Resources. Edited by Fuller W.A., Kevan P.G. Morges, Switzerland. pp. 85–92.

Beven K., Kirkby M. 1979. A physically based, variable contributing area model of basin hydrology/Un modèle à base physique de zone d'appel variable de l'hydrologie du bassin versant. Hydrological Sciences Bulletin, 24: 43–69.

Bintanja R. 2018. The impact of Arctic warming on increased rainfall. Scientific Reports, 8: 16001.

Bjorkman A.D., Vellend M., Frei E.R., Henry G.H.R. 2017. Climate adaptation is not enough: warming does not facilitate success of southern tundra plant populations in the high Arctic. Global Change Biology, 23(4): 1540–1551.

Bliss L.C., Svoboda J. 1984. Plant communities and plant production in the western Queen Elizabeth Islands. Holarctic Ecology, 7: 325–344.

Bliss L.C., Svoboda J., Bliss D.I. 1984. Polar deserts, their plant cover and plant production in the Canadian High Arctic. Holarctic Ecology, 7: 305–324.

Bliss L.C., Matveyeva N.V. 1992. Circumpolar Arctic vegetation. In Arctic ecosystems in a Changing Climate—an ecophysiological perspective. Edited by Chapin S.F., Jeffries R.L., Reynolds J.F., Shaver G.R., Svoboda J., Chu E.W. Academic Press Inc., ​​​​​​ San Diego, CA. pp. 59–89.

Boelman N.T., Stieglitz M., Rueth H.M., Sommerkorn M., Griffin K.L., Shaver G.R., Gamon J.A. 2003. Response of NDVI, biomass, and ecosystem gas exchange to long-term warming and fertilization in wet sedge tundra. Oecologia, 135: 414–421.

Bolduc C., Lamoureux S.F., Franssen J. 2018. Thermal and isotopic evidence for surface and subsurface water contributions to baseflow in a high Arctic river. Hydrological Processes, 32: 602–616.

Bonney M.T., Danby R.K., Treitz P.M. 2018. Landscape variability of vegetation change across the forest to tundra transition of central Canada. Remote Sensing of Environment, 217: 18–29.

Box J.E., Colgan W.T., Christensen T.R., Schmidt N.M., Lund M., Parmentier F.-J., et al. 2019. Key indicators of Arctic climate change: 1971-2017. Environmental Research Letters, 14: 045010.

Brassard G.R. 1967. A contribution to the bryology of Melville Island, N.W.T. The Bryologist, American Bryological and Lichenological Society, 20(3): 347–351.

Braun-Blanquet J. 1932. Plant sociology: the study of plant communities. 1st ed. McGraw-Hill, New York, NY.

Braybrook C.A., Scott N.A., Treitz P.M., Humphreys E.R. 2021. Inter-annual variability of summer net ecosystem CO2 exchange in High Arctic tundra. Journal of Geophysical Research: Biogeosciences, 126: e2020JG006.

Burn C.R. 2000. The thermal regime of a retrogressive thaw slump near Mayo, Yukon Territory. Canadian Journal of Earth Sciences, 37: 967–981.

CAVM Team. 2003. Circumpolar Arctic vegetation map (1:7,500,000 scale). Conservation of Arctic Flora and Fauna (CAFF) Map No. 1. U.S. Fish and Wildlife Service, Anchorage, AL.

CAVM Team. 2023. Raster circumpolar Arctic vegetation map, version 2. Scale 1:7,000,000. Mendeley Data.

Celis G., Mauritz M., Bracho R., Salmon V.G., Webb E.E., Hutchings J., et al. 2017. Tundra is a consistent source of CO2 at a site with progressive permafrost thaw during 6 years of chamber and eddy covariance measurements. Journal of Geophysical Research: Biogeosciences, 122: 1471–1485.

Chadburn S.E., Burke E.J., Cos P.M., Friedlingstein P., Hugelius G., Westermann S. 2017. An observation-based constraint on permafrost loss as a function of global warming. Nature Climate Change, 7: 340–344.

Chapin F.S. III, Sturm M., Serreze M.C., McFadden J.P., Key J.R., Lloyd A.H., et al. 2005. Role of land-surface changes in Arctic summer warming. Science, 310(5748): 657–660.

Chylek P., Folland C., Klett J. D., Wang M., Hengartner N., Lesins G., Dubey M. K. 2022. Annual mean Arctic amplification 1970–2020: observed and simulated by CMIP6 climate models.Geophysical Research Letters, 49: e2022GL099371.

Cockburn J.M.H., Lamoureux S.F. 2008. Hydroclimate controls over seasonal sediment yield in two adjacent High Arctic watersheds. Hydrological Processes, 22: 2013–2027.

Collingwood A. 2014. Modelling biophysical variables in the Canadian High Arctic using synthetic aperture radar data. Ph.D. thesis, Queen's University, Kingston, ON, Canada.

Collingwood A., Treitz P.M., Charbonneau F., Atkinson D.M. 2014a. Artificial neural network modeling of high Arctic phytomass using synthetic aperture radar and multispectral data. Remote Sensing, 6: 2134–2153.

Collingwood A., Treitz P.M., Charbonneau F. 2014b. Surface roughness estimation from RADARSAT-2 data in a High Arctic environment. International Journal of Applied Earth Observation and Geoinformation, 27: 70–80.

Collingwood A., Charbonneau F., Shang C., Treitz P.M. 2018. Spatiotemporal variability of Arctic soil moisture detected from high resolution RADARSAT-2 SAR data. Advances in Meteorology, 2018: 17.

Commanea R., Lindaas J., Benmergui J., Luus K.A., Chang R.Y.-W., Daube B.C., Euskirchen E.S., et al. 2017. Carbon dioxide sources from Alaska driven by increasing early winter respiration from Arctic tundra. Proceedings of the National Academy of Sciences of the United States of America, 114(21): 5361–5366.

Crosson W.L., Laymon C.A., Inguva R., Schamschula M.P. 2002. Assimilating remote sensing data in a surface flux–soil moisture model. Hydrological Processes, 16: 1645–1662.

Cuven S., Francus P., Lamoureux S.F. 2010. Estimation of grain size variability with micro X-ray fluorescence in laminated lacustrine sediments, Cape Bounty, Canadian High Arctic. Journal of Paleolimnology, 44: 803–817.

Cuven S., Francus P., Lamoureux S.F. 2011. Mid to Late Holocene hydroclimatic and geochemical records from the varved sediments of East Lake, Cape Bounty, Canadian High Arctic. Quaternary Science Reviews, 30: 2651–2665.

Dagg J., Lafleur P. 2011. Vegetation community, foliar nitrogen, and temperature effects on tundra CO2 exchange across a soil moisture gradient. Arctic, Antarctic, and Alpine Research, 43: 189–197.

Desjardins É., Lai S., Houle L., Caron A., Thériault V., Tam A., et al. 2023. Algorithms and predictors for land cover classification of polar deserts: a case study highlighting challenges and recommendations for future applications. Remote Sensing, 15(12): 3090.

Diepstraten R.A.E., Jessen T.D., Fauvelle C.M.D., Musiani M.M. 2018. Does climate change and plant phenology research neglect the Arctic tundra? Ecosphere, 9(9): e02362.

Dugan H.A., Lamoureux S.F., Lewis T., Lafrenière M.J. 2012. The impact of permafrost disturbances and sediment loading on the limnological characteristics of two High Arctic lakes. Permafrost and Periglacial Processes, 23: 119–126.

Ecosystem Classification Group. 2008. Ecological regions of the Northwest Territories—Taiga Shield. Department of Environment and Natural Resources, Government of the Northwest Territories, Yellowknife, NT, Canada.

Ecosystem Classification Group. 2012. Ecological regions of the Northwest Territories—Southern Arctic. Department of Environment and Natural Resources, Government of the Northwest Territories, Yellowknife, NT, Canada.

Edlund S.A. 1983. The distribution of plant communities on Melville Island, Arctic Canada. In The geology of Melville Island, Arctic Canada. Edited by Christie R.L., McMillan N.J. Geological Survey of Canada, Bulletin 450., pp. 247–255.

Edlund S.A., Alt B.T. 1989. Regional congruence of vegetation and summer climate patterns in the Queen Elizabeth Islands, Northwest Territories, Canada. Arctic, 42: 3–23.

Edwards R., Treitz P.M. 2017. Vegetation greening trends at two sites in the Canadian Arctic: 1984–2015. Arctic, Antarctic, and Alpine Research, 49(4): 601–619.

Elmendorf S., Henry G., Hollister R., Björk R.G., Boulanger-Lapointe N., Cooper E.J., Cornelissen J.H.C., et al. 2012. Plot-scale evidence of tundra vegetation change and links to recent summer warming. Nature Climate Change, 2: 453–457.

Emmerton C.A., St. Louis V.L., Humphreys E.R., Gamon J.A., Barker J.D., Pastorello G.Z. 2016. Net ecosystem exchange of CO2 with rapidly changing high Arctic landscapes. Global Change Biology, 22: 1185–1200.

England J.H., Furze M.F.A., Doupe J.P. 2009. Revision of the NW Laurentide Ice Sheet: implications for paleoclimate, in the north–east extremity of Beringia, and Arctic Ocean sedimentation. Quaternary Science Reviews, 28: 1573–1596.

Epstein H.E., Raynolds M.K., Walker D.A., Bhatt U.S., Tucker C.J., Pinzon J.E. 2012. Dynamics of aboveground phytomass of the circumpolar Arctic tundra during the past three decades. Environmental Research Letters, 7(1): 015506.

Epstein H.E., Myers-Smith I., Walker D.A. 2013. Recent dynamics of arctic and sub-arctic vegetation. Environmental Research Letters, 8(1): 015040.

Evans B.M., Walker D.A., Benson C.S., Nordstrand E.A., Petersen G.W. 1989. Spatial interrelationships between terrain, snow distribution and vegetation patterns at an arctic foothills site in Alaska. Ecography, 12(3): 270–278.

Favaro E.A., Lamoureux S.F. 2014. Antecedent controls on rainfall runoff response and sediment transport in a High Arctic catchment. Geografiska Annaler: Series A, Physical Geography, 96: 433–446.

Favaro E.A., Lamoureux S.F. 2015. Downstream patterns of suspended sediment transport in a High Arctic river influenced by permafrost disturbance and recent climate change. Geomorphology, 246: 359–369.

Fouché J., Lafrenière M.J., Rutherford K., Lamoureux S.F. 2017. Seasonal hydrology and permafrost disturbance impacts on dissolved organic matter composition in High Arctic headwater catchments. Arctic Science, 3: 378–405.

Fouché J., Christiansen C.T., Lafrenière M.J., Grogan P., Lamoureux S.F. 2020. Seasonal hydrology and permafrost disturbance impacts on dissolved organic matter composition in High Arctic headwater catchments. Nature Communications, 11: 4500.

Fraser R., Olthof I., Lantz T.C., Schmitt C. 2016. UAV photogrammetry for mapping vegetation in the low-Arctic. Arctic Science, 2: 79–102.

Freemantle V., Freemantle J., Atkinson D., Treitz P.M. 2020. A high spatial resolution satellite remote sensing time series analysis of Cape Bounty, Melville Island, Nunavut (2004–2018). Canadian Journal of Remote Sensing, 46(6): 733–752.

Fu P., Rich P. 1999. Design and implementation of the Solar Analyst: an ArcView extension for modeling solar radiation at landscape scales. In Proceedings of the Nineteenth Annual ESRI User Conference, San Diego. Available from http://professorpaul.com/publications/fu_rich_1999_esri.pdf.

García Criado M., Myers-Smith I., Bjorkman A., Elmendorf S., Normand S., Aastrup P., et al. 2023. Plant diversity dynamics over space and time in a warming Arctic. California Digital Library (CDL).

Gaspard A., Simard M., Boudreau S. 2023. Patterns and drivers of change in the normalized difference vegetation index in Nunavik (Québec, Canada) over the period 1984-2020. Atmosphere, 14: 1115.

Gaussman H.W., Allen W.A., Cardenas R. 1969. Reflectance of cotton leaves and their structure. Remote Sensing of Environment, 1: 110–122.

Goosse H., Kay J.E., Armour K.C., Bodas-Salcedo A., Chepfer H., Docquier D., et al. 2018. Quantifying climate feedbacks in polar regions. Nature Communications, 9: 1919.

Grayson R., Blöschl G. (Editors). 2001. Spatial patterns in catchment hydrology—observations and modelling. Cambridge University Press, Cambridge, UK. 416 p.

Gregory F.M. 2011. Biophysical remote sensing and terrestrial CO2 exchange at Cape Bounty, Melville Island. M.Sc. thesis, Queen's University, Kingston, ON.

Griggs R.F. 1934. The problem of arctic vegetation. Journal of the Washington Academy of Sciences, 24(4): 153–175.

Hadjimitisis D.G., Papadavid G., Agapiou A., Themistocleous K., Hadjimitisis M.G., Retalis A., et al. 2010. Atmospheric correction for satellite remotely sensed data intended for agricultural applications: impact on vegetation indices. Natural Hazards Earth Systems Science, 10: 89–96.

Heijmans M.M.P.D., Magnússon R.Í., Lara M.J., Frost G.V., Myers-Smith I.H., Huissteden J. van, et al. 2022. Tundra vegetation change and impacts on permafrost. Nature Reviews—Earth and Environment, 3: 68–84.

Hinkel K.M., Paetzold F., Nelson F.E., Bockheim J.G. 2001. Patterns of soil temperature and moisture in the active layer and upper permafrost at Barrow, Alaska: 1993–1999. Global and Planetary Change, 29(3–4): 293–309.

Hobbie J.E., Shaver G.R., Rastetter E.B., Cherry J.E., Goetz S.J., Guay K.C., et al. 2017. Ecosystem responses to climate change at a Low Arctic and a High Arctic long-term research site. Ambio, 46(Suppl. 1): 160–173.

Hodgson D.A., Vincent J.-S., Fyles J.G. 1984. Quaternary geology of central Melville Island, Northwest Territories, Geological Survey of Canada. Paper 83-16, Map 1583A. p. 25

Holloway J.E., Lamoureux S.F., Montross S.N., Lafrenière M.J. 2016. Climate and terrain characteristics linked to mud ejection occurrence in the Canadian High Arctic. Permafrost and Periglacial Processes, 27(2): 204–217.

Holloway J.E., Rudy A.C.A., Lamoureux S.F., Treitz P.M. 2017. Determining the terrain characteristics related to surface expression of subsurface water pressurization in permafrost landscapes using susceptibility modelling. The Cryosphere, 11: 1403–1415.

Hugelius G., Strauss J., Zubrzycki S., Harden J.W., Schuur E.A.G., Ping C.-L., et al. 2014. Estimated stocks of circumpolar permafrost carbon with quantified uncertainty ranges and identified data gaps. Biogeosciences, 11: 6573–6593.

Hung J.K.Y., Treitz P.M. 2020. Environmental land-cover classification for integrated watershed studies: Cape Bounty, Melville Island, Nunavut. Arctic Science, 6: 404–422.

Hung J.K.Y., Scott N.A., Treitz P.M. 2023. Investigating ten years of warming and enhanced snow depth on nutrient availability and greenhouse gas fluxes in a High Arctic ecosystem. Arctic, Antarctic, and Alpine Research, 55(1): 2178428.

Illeris L., Michelsen A., Jonasson S. 2003. Soil plus root respiration and microbial biomass following water, nitrogen, and phosphorus application at a high arctic semi desert. Biogeochemistry, 65: 15–29.

IPCC. 2021. Climate Change 2021: the physical science basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Edited by Masson-Delmotte V., Zhai P., Pirani A., Connors S.L., Péan C., Berger, S. et al. et al. Cambridge University Press, Cambridge, New York, NY.

Isaksen K., Nordli Ø., Ivanov B., Koltzow M.A.O., Aaboe S., Gjelten H.M., et al. 2022. Exceptional warming over the Barents area. Scientific Reports, 12: 9371.

Jenness J. 2006. Topographic position index (tpi_jen.avx) extension for ArcView 3.x, v.1.2. Jenness Enterprises. Available from http://www.jennessent.com/arcview/tpi.htm.

Jensen J.R. 2016. Introductory digital image processing: a remote sensing perspective. 4th ed. Pearson Inc. 623p.

Jones B.M., Schaeffer Tessier S., Tessier T., Brubaker M., Brook M., Schaeffer J., et al. 2023. Integrating local environmental observations and remote sensing to better understand the life cycle of a thermokarst lake in Arctic Alaska. Arctic, Antarctic, and Alpine Research, 55(1): 2195518.

Karlson M., Bastviken D., Reese H. 2021. Error characteristics of Pan-Arctic digital elevation models and elevation derivatives in northern Sweden. Remote Sensing, 13: 4653.

Kokelj S.V., Jorgenson M.T. 2013. Advances in thermokarst research. Permafrost & Periglacial, 24: 108–119.

Koven C.D., Ringeval B., Friedlingstein P., Ciais P., Cadule P., Khvorostyanov D., et al. 2011. Permafrost carbon-climate feedbacks accelerate global warming. Proceedings of the National Academy of Sciences of the United States of America, 108: 14769–14774.

Lafarge-England C. 1989. The contemporary moss assemblages of a high arctic upland, northern Ellesmere Island, N.W.T. Canada. Canadian Journal of Botany, 67: 491–504.

Lafleur P.M., Humphreys E.R. 2018. Tundra shrub effects on growing season energy and carbon dioxide exchange. Environmental Research Letters, 13: 055001.

Lafrenière M.J., Lamoureux S.F. 2013. Thermal perturbation and rainfall runoff have greater impact on seasonal solute loads than physical disturbance of the active layer. Permafrost and Periglacial Processes, 24: 241–251.

Lajeunesse P., Hanson M.A. 2008. Field observations of recent transgression on northern and eastern Melville Island, western Canadian Arctic Archipelago, Geomorphology, 101: 618–630.

Laidler G., Treitz P., Atkinson D. 2009. Remote sensing of Arctic vegetation: relations between the NDVI, spatial resolution and vegetation cover on Boothia Peninsula, Nunavut. Arctic, 61(1): 1–13.

Lamoureux S.F., Lafrenière M.J. 2009. Fluvial impact of extensive active layer detachments, Cape Bounty, Melville Island, Canada. Arctic, Antarctic, and Alpine Research, 41: 59–68.

Lamoureux S.F., Lafrenière M.J. 2017. More than just snowmelt: integrated watershed science for changing climate and permafrost at the Cape Bounty Arctic Watershed Observatory. WIREs Water, 5(1): e1255.

Langford Z.L., Kumar J., Hoffman F.M., Breen A.L., Iversen C.M. 2019. Arctic vegetation mapping using unsupervised training datasets and convolution neural networks. Remote Sensing, 11(1): 69.

Laudon H., Spence C., Buttle J., Carey S.K., McDonnell J.J., McNamara J.P. 2017. Save northern high-latitude catchments. Nature Geoscience, 10: 324–325.

Lewis T., Lafrenière M.J., Lamoureux S.F. 2012. Hydrochemical and sedimentary responses of paired High Arctic watersheds to unusual climate and permafrost disturbance, Cape Bounty, Melville Island, Canada. Hydrological Processes, 26(13): 2003–2018.

Lewkowicz A.G. 2007. Dynamics of active-layer detachment failures, Fosheim peninsula, Ellesmere Island, Nunavut, Canada. Permafrost & Periglacial, 18: 89–103.

Likens G.E. 2001. Biogeochemistry, the watershed approach: some uses and limitations. Marine Freshwater Research, 52: 5–12.

Liu N., Treitz P.M. 2016. Modelling High Arctic percent vegetation cover using field digital images and high-resolution satellite data. International Journal of Applied Earth Observation and Geoinformation, 52: 445–456.

Liu N., Budkewitsch P., Treitz P.M. 2017. Examining spectral reflectance features related to Arctic percent vegetation cover: implications for hyperspectral remote sensing of Arctic tundra.Remote Sensing of Environment, 192: 58–72.

Liu N., Treitz P.M. 2018. Remote sensing of Arctic percent vegetation cover and fAPAR on Baffin Island, Nunavut, Canada. International Journal of Applied Earth Observation and Geoinformation, 71: 159–169.

Macander M.J., Frost G.V., Nelson P.R., Swingley C.S. 2017. Regional quantitative cover mapping of tundra plant functional types in Arctic Alaska. Remote Sensing, 9(10): 1024.

Maher A., Treitz P., Ferguson M. 2012. Can Landsat data detect variations in snow cover within habitats of arctic ungulates? Wildlife Biology, 18: 75.

Mallory M.L., Gilchrist H.G., Janssen M., Major H.L., Merkel F., Provencher J.F., Strøm H. 2018. Financial costs of conducting science in the Arctic: examples from seabird research. Arctic Science, 4: 624–633.

McCrystall M.R., Stroeve J., Serreze M., Forbes B.C., Screen J.A. 2021. New climate models reveal faster and larger increases in Artic precipitation than previously projected. Nature Communications, 12: 6765.

McFadden S. 2019. Fine-scale ground surface vertical displacement and soil water processes in the Canadian High Arctic. M.Sc. thesis, Department of Geography and Planning, Queen's University, Kingston, ON, Canada

McLennan D.S., MacKenzie W.H., Meidinger D., Wagner J., Arko C. 2018. A standardized ecosystem classification for the coordination and design of long-term terrestrial ecosystem monitoring in Arctic-Subarctic biome. Arctic, 71(5): 1–15.

Myers-Smith I.H., Forbes B.C., Wilmking M., Hallinger M., Lantz T., Blok D., et al. 2011. Shrub expansion in tundra ecosystems: dynamics, impacts and research priorities. Environmental Research Letters, 6: 045509.

Mikola J., Virtanen T., Linkosalmi M., Vähä E., Nyman J., Postanogova O., et al. 2018. Spatial variation and linkages of soil and vegetation in the Siberian Arctic tundra—coupling field observations with remote sensing data. Biogeosciences, 15(9): 2781–2801.

Miner K.R., Turetsky M.R., Malina E., Bartsch A., Tamminen J., McGuire A.D., et al. 2022. Permafrost carbon emissions in a changing Arctic. Nature Reviews Earth & Environment, 3: 55–67.

Natali S.M., Schuur E.A., Rubin R.L. 2012. Increased plant productivity in Alaskan tundra as a result of experimental warming of soil and permafrost. Journal of Ecology, 100: 488–498.

Natali S.M., Watts J.D., Rogers B.M., Potter S. , Ludwig S.M., Selbmann A-K., et al. 2019. Large loss of CO2 in winter observed across the northern permafrost region. Nature Climate Change, 9: 852–857.

Nobrega S., Grogan P. 2008. Landscape and ecosystem-level controls on net carbon dioxide exchange along a natural moisture gradient in Canadian Low Arctic tundra. Ecosystems, 11: 377–396.

Normandeau A., Lamoureux S.F., Lajeunesse P., Francus P. 2016. Sediment dynamics in paired High Arctic lakes revealed from high-resolution swath bathymetry and acoustic stratigraphy surveys. Journal of Geophysical Research: Earth Surface, 121: 1676–1696.

Oberbauer S.F., Tweedie C.E., Welker J.M., Fahnestock J.T., Henry G.H.R., Webber P.J., et al. 2007. Tundra CO2 fluxes in response to experimental warming across latitudinal and moisture gradients. Ecological Monographs, 77(2): 221–238.

Ogden E.L., Cumming S.G., Smith S.L., Turetsky M.R., Baltzer J.L. 2023. Permafrost thaw induces short-term increase in vegetation productivity in northwestern Canada. Global Change Biology, 2023: 1–15.

O'Neill H.B., Wolfe S.A., Duchesne C. 2019. New ground ice maps for Canada using a paleogeographic modeling approach. The Cryosphere, 13: 753–773.

Ono J., Watanabe M., Komuro Y., Tatebe H., Abe M. 2022. Enhanced Arctic warming amplification revealed in a low-emission scenario. Nature Communications: Earth and Environment, 3(27).

Parker G.R., Ross R.K. 1975. Summer habitat use by muskoxen (Ovibos moschatus) and peary caribou (Rangifer tarandus pearyi) in the Canadian High Arctic. Polarforschung, 1: 12–25.

Pavelka M., Acosta M., Kiese R., Altimir N., Brümmer C., Crill P., et al. 2018. Standardisation of chamber technique for CO2, N2O and CH4 fluxes measurements from terrestrial ecosystems. International Agrophysics, 32: 569–587.

Plaza C., Pegoraro E., Bracho R., Celis G., Crummer K.G., Hutchings J.A., et al. 2019. Direct observation of permafrost degradation and rapid soil carbon loss in tundra. Nature Geoscience, 12: 627–631.

Ponomarenko S., McLennan D., Pouliot D.A., Wagner J. 2019. High resolution mapping of tundra ecosystems on Victoria Island, Nunavut—application of a standardized terrestrial ecosystem classification. Canadian Journal of Remote Sensing, 45(5): 551–571.

Porter C., Morin P., Howat I., Noh M.-J., Bates B., Peterman K., et al. 2018. ArcticDEM version 3, Harvard dataverse. Polar Geospatial Centre, University of Minnesota, Saint Paul, MN.

Post E., Alley R.B., Christensen T.R., Macias-Fauria M., Forbes B.C., Gooseff M.N., et al. 2019. The polar regions in a 2° C warmer world. Science Advances, 5(12): eaaw9883.

Previdi M., Smith K., Polvani L.M. 2021. Arctic amplification of climate change: a review of underlying mechanisms. Environmental Research Letters, 16: 093003.

Purevdorj T.S., Tateishi R., Ishiyama T., Honda Y. 1998. Relationships between percent vegetation cover and vegetation indices. International Journal of Remote Sensing, 19(18): 3519–3535.

Rantanen M., Karpechko A.Yu., Lipponen A., Nordling K., Hyvärinen O., Ruosteenoja K., et al. 2022. The Arctic has warmed nearly four times faster than the globe since 1979. Nature Communications—Earth and Environment, 3: 168.

Raynolds M.K., Walker D.A., Maier H.A. 2006. NDVI patterns and phytomass distribution in the circumpolar Arctic. Remote Sensing of Environment, 102: 271–281.

Raynolds M.K., Walker D.A., Epstein H.E., Pinzon J.E., Tucker C.J. 2012. A new estimate of tundra-biome phytomass from trans-Arctic field data and AVHRR NDVI. Remote Sensing Letters, 3(5): 403–411.

Raynolds M.K., Walker D.A., Balser A., Bay C., Campbell M., Cherosov M.M., et al. 2019. A raster version of the Circumpolar Arctic Vegetation Map (CAVM). Remote Sensing of Environment, 232: 111297.

Reichle L.M., Epstein H.E., Bhatt U.S., Raynolds M.K., Walker D.A. 2018. Spatial heterogeneity of the temporal dynamics of Arctic tundra vegetation. Geophysical Research Letters, 45(17): 9206–9215.

Reichstein M., Bahn M., Ciais P., Frank D., Mahecha M. D., Seneviratne S. I., et al. 2013. Climate extremes and the carbon cycle. Nature, 500(7462): 287–295.

Richardson A.D., Hollinger D.Y., Shoemaker J.K., Hughes H., Savage K., Davidson E.A. 2019. Six years of ecosystem-atmosphere greenhouse gas fluxes measured in a sub-boreal forest. Scientific Data, 6: 117.

Ritchie J.C. 1993. Northern vegetation. In Canada's cold environments. Edited by French H.M., Slaymaker O. McGill-Queen's Press. pp. 93–116.

Roberts K.E., Lamoureux S.F., Kyser T.K., Muir D.C.G., Lafrenière M.J., Iqaluk D., et al. 2017. Climate and permafrost effects on the chemistry and ecosystems of High Arctic Lakes. Scientific Reports, 7: 13292.

Robson G., Treitz P.M., Lamoureux S.F., Murnaghan K., Brisco B. 2021. Seasonal surface subsidence and frost heave detected by C-Band DInSAR in a High Arctic environment, Cape Bounty, Melville Island, Nunavut, Canada. Remote Sensing, 13(2505).

Rouse J.W., Haas R.H., Schell J.A., Deering D.W. 1973. Monitoring vegetation systems in the Great Plains with ERTS. In Third Earth Resources Technology Satellite-1 Symposium. Proceedings. Edited by Freden S.C., Mercanti E.P., Becker M.A. Conference held 10–14 December 1973 at the National Aeronautics and Space Administration, Scientific and Technical Information Office, Goddard Space Flight Center, Washington, DC. pp. 309–317.

Rudd D.A., Karami M., Fensholt R. 2021. Towards high-resolution land-cover classification of Greenland: a case study covering Kobbefjord, Disko and Zackenberg. Remote Sensing, 13: 3559.

Rudy A.C.A., Lamoureux S., Treitz P.M., Collingwood A. 2013. Identifying permafrost slope disturbance using multi-temporal optical satellite images and change detection techniques. Cold Regions Science and Technology, 88: 37–49.

Rudy A.C.A., Lamoureux S.F., Treitz P.M., Ewijk van K. 2016. Transferability of regional permafrost disturbance susceptibility modelling using generalized linear and generalized additive models. Geomorphology, 264: 95–108.

Rudy A.C.A., Lamoureux S.F., Treitz P.M., Ewijk van K., Bonnaventure P., Budkewitsch P. 2017. Terrain controls and landscape-scale modelling of active-layer detachments, Sabine Peninsula, Melville Island, Nunavut. Permafrost and Periglacial Processes, 28: 79–91.

Rudy A.C.A., Lamoureux S.F., Treitz P.M., Short N., Brisco B. 2018. Seasonal and multi-year surface displacements measured by DInSAR in a high Arctic permafrost environment. International Journal of Applied Earth Observation and Geoinformation, 64: 51–61.

Schuur E.A.G., McGuire A.D., Schädel C., Grosse G., Harden J.W., Hayes D.J., et al. 2015. Climate change and the permafrost carbon feedback. Nature, 520: 171–179.

Screen J.A., Simmonds I. 2010. The central role of diminishing sea ice in recent Arctic temperature amplification. Nature, 464(7293): 1334–1337.

Sellers H.L., Vargas Zesati S.A., Elmendorf S.C., Locher A., Oberbauer S.F., Tweedie C.E., et al. 2023. Can plot-level photographs accurately estimate tundra vegetation cover in northern Alaska? Remote Sensing, 15(8).

Shaver G.R., Street L.E., Rastetter E.B., van Wijk M.T., Williams M. 2007. Functional convergence in regulation of net CO2 flux in heterogeneous tundra landscapes in Alaska and Sweden. Journal of Ecology, 95: 802–817.

Shiklomanov N.I., Streletskiy D.A., Little J.D., Nelson F.E. 2013. Isotropic thaw subsidence in undisturbed permafrost landscapes. Geophysical Research Letters, 40: 6356–6361.

Short N.H., Fraser R.H. 2023. Comparison of RADARSAT-2 and Sentinel-1 DInSAR displacements over upland ice-wedge polygonal terrain, Banks Island, Northwest Territories, Canada. Geomatics Canada Technical Report, Open File 73. 22 p.

Shu Q., Wang Q., Årthun M., Wang S., Song Z., Zhang M., Qiao F. 2022. Arctic Ocean amplification in a warming climate in CMIP6 models. Science Advances, 8: eabn9755.

Sluijs van der J., Saiet E., Bakelaar C.N., Wentworth A., Fraser R.H., Kokelj S.V. 2023. Beyond visual-line-of-sight (BVLOS) drone operations for environmental and infrastructure monitoring: a case study in northwestern Canada. Drone Systems and Applications, 11: 1–15.

Stanski K., Myers-Smith I.H., Lucas C.G. 2021. Flower detection using object analysis: new ways to quantify plant phenology in a warming tundra biome. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 14: 9287–9296.

Stewart K.A., Lamoureux S.F. 2011. Seasonal hydrochemical conditions and limnological response in adjacent high Arctic lakes: Cape Bounty, Melville Island, Nunavut. Arctic, 64: 169–182.

Stewart K.A., Lamoureux S.F. 2012. Seasonal and microhabitat influences on diatom assemblages and their representation in sediment traps and surface sediments from adjacent High Arctic lakes: Cape Bounty, Melville Island, Nunavut. Hydrobiologia, 683: 265–286.

Timoney K.P., La Roi G.H., Zoltai S.C., Robinson A.L. 1992. The high Subarctic forest-tundra of northwestern Canada: position, width, and vegetation gradients in relation to climate. Arctic, 45(1): 1–9.

Thomson E.R., Spiegel M.P., Althuizen I.H.J., Bass P., Chen S., Chmurzynski A., et al. 2021. Multiscale mapping of plant functional groups and plant traits in the High Arctic using field spectroscopy, UAV imagery and Sentinel-2A data. Environmental Research Letters, 16: 055006.

Toth C., Jóźków G. 2016. Remote sensing platforms and sensors: a survey. ISPRS Journal of Photogrammetry and Remote Sensing, 115: 22–36.

Treitz P., Thomas V., Zarco-Tejada P., Gong P., Curran P. 2010. Hyperspectral remote sensing for forestry. Monograph Series. American Society for Photogrammetry and Remote Sensing, Bethesda, MD. 107 p.

Tucker C.J. 1979. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sensing of Environment, 8(2): 127–150.

Virkkala A.-M., Virtanen T., Lehtonen A., Rinne J., Luoto M. 2018. The current state of CO2 flux chamber studies in the Arctic tundra: a review. Progress in Physical Geography: Earth and Environment, 42: 162–184.

Virkkala A.-M., Niittynen P., Kemppinen J., Marushchak M.E., Voigt C., Hensgens G., et al. 2023. High resolution spatial patterns and drivers of terrestrial ecosystem carbon dioxide, methane, and nitrous oxide fluxes in the tundra. Biogeosciences, Discussions(preprint).

Virtanen T., Ek M. 2014. The fragmented nature of tundra landscape. International Journal of Applied Earth Observation and Geoinformation, 27: 4–12.

Wagner I., Hung J.K.Y., Neil A., Scott N.A. 2019. Net greenhouse gas fluxes from three High Arctic plant communities along a moisture gradient. Arctic Science, 5(4): 185–201.

Walker D.A., Raynolds M.K., Daniels F.J.A., Einarsson E., Elvebakk A., Gould W.A., et al. 2005. The circumpolar Arctic vegetation map. Journal of Vegetation Science, 16: 267–282.

Walker D.A., Kuss P., Epstein H.E., Kade A.N., Vonlanthen C.M., Raynolds M.K., Daniels F.J.A. 2011. Vegetation of zonal patterned-ground ecosystems along the North American Arctic Transect. Applied Vegetation Science, 14(4): 440–463.

Walker D.A., Epstein H.E., Raynolds M.K., Kuss P., Kopecky M.A., Frost G.V., et al. 2012. Environment, vegetation and greenness (NDVI) along the North America and Eurasia Arctic transects. Environmental Research Letters, 7(1).

Walker D.A., Daniëls F.J.A., Alsos I., Bhatt U.S., Breen A.L., Buchhorn M., et al. 2016. Circumpolar Arctic vegetation: a hierarchic review and roadmap toward an internationally consistent approach to survey, archive and classify tundra plot data. Environmental Research Letters, 11(5): 055005.

Walker D.A., Daniëls F.J.A., Matveyeva N.V., Šibík J., Walker M.D., Breen A.L., et al. 2018. Circumpolar Arctic vegetation classification. Phytocoenolgica, 48(2): 181–201.

Wall J., Collingwood A., Treitz P. 2010. Monitoring surface moisture state in the Canadian High Arctic using synthetic aperture radar (SAR). Canadian Journal of Remote Sensing, 36: S124–S134.

Wang. A., Chlus A., Geygan R., Zhiwei Y., Zheng T., Singh A., et al. 2020. Folia functional traits from imaging spectroscopy across biomes in eastern North America. New Phytologist, 228: 494–511.

Watts J.D., Natali S.M., Minions C., Risk D., Arndt K., Zona D., et al. 2021. Soil respiration strongly offsets carbon uptake in Alaska and Northwest Canada. Environmental Research Letters, 16: 084051.

Wijk van M.T., Williams M. 2005. Optical instruments for measuring leaf area index in low vegetation: application in Arctic ecosystems. Ecological Applications, 15(4): 1462–1470.

Wilson J.P., Gallant J.C. 2000. Digital terrain analysis. In Terrain analysis: principles and applications. Edited by Wilson J.P., Gallant J.C. 1st ed. John Wiley and Sons, New York, NY. pp. 1–27.

Wolfe S.A., Short N.H., Morse P.D., Schwarz S.H., Stevens C.W. 2014. Evaluation of RADARSAT-2 DInSAR seasonal surface displacement in discontinuous permafrost terrain, Yellowknife, Northwest Territories, Canada. Canadian Journal of Remote Sensing, 40: 406–422.

Wolff F., Kolari T.H.M., Villoslada M., Tahvanainen T., Korpelainen P., Zamboni P.A., Kumpula T. 2023. RGB vs. multispectral imagery: mapping aapa mire plant communities with UAVs. Ecological Indicators, 148: 110140.

Woo M.-K., Xia Z. 1995. Suprapermafrost groundwater seepage in gravelly terrain, Resolute, NWT, Canada. Permafrost & Periglacial, 6(1): 57–72.

Wookey P.A., Aerts R., Bardgett R.D., Baptist F., Bråthen K.A., Cornelissen J.H.C., et al. 2009. Ecosystem feedbacks and cascade processes: understanding their role in the responses of Arctic and Alpine ecosystems to environmental change. Global Change Biology, 15(5): 1153–1172.

Wright C.M., Blaser A., Treitz P.M., Scott N.A. 2021. Spatial variability in carbon exchange processes within wet sedge meadows in the Canadian High Arctic. Advances in Polar Science, 32(1): 1–19.

Zhang B., Wu Y., Zhao B., Chanussot J., Hong D., Yao J., Gao L. 2022. Progress and challenges in intelligent remote sensing satellite systems. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 15: 1814–1822.