Search for: All records

The Arctic is warming faster than anywhere else on Earth, placing tundra ecosystems at the forefront of global climate change. Plant biomass is a fundamental ecosystem attribute that is sensitive to changes in climate, closely tied to ecological function, and crucial for constraining ecosystem carbon dynamics. However, the amount, functional composition, and distribution of plant biomass are only coarsely quantified across the Arctic. Therefore, we developed the first moderate resolution (30 m) maps of live aboveground plant biomass (g m− 2) and woody plant dominance (%) for the Arctic tundra biome, including the mountainous Oro Arctic. We modeled biomass for the year 2020 using a new synthesis dataset of field biomass harvest measurements, Landsat satellite seasonal synthetic composites, ancillary geospatial data, and machine learning models. Additionally, we quantified pixel-wise uncertainty in biomass predictions using Monte Carlo simulations and validated the models using a robust, spatially blocked and nested cross-validation procedure. Observed plant and woody plant biomass values ranged from 0 to ~6000 g m− 2 (mean ≈350 g m− 2), while predicted values ranged from 0 to ~4000 g m− 2 (mean ≈275 g m− 2), resulting in model validation root-mean-squared-error (RMSE) ≈400 g m− 2 and R2 ≈ 0.6. Our maps not only capture large-scale patterns of plant biomass and woody plant dominance across the Arctic that are linked to climatic variation (e.g., thawing degree days), but also illustrate how fine-scale patterns are shaped by local surface hydrology, topography, and past disturbance. By providing data on plant biomass across Arctic tundra ecosystems at the highest resolution to date, our maps can significantly advance research and inform decision-making on topics ranging from Arctic vegetation monitoring and wildlife conservation to carbon accounting and land surface modeling 

Abstract Climate change is driving substantial changes in North American boreal forests, including changes in productivity, mortality, recruitment, and biomass. Despite the importance for carbon budgets and informing management decisions, there is a lack of near‐term (5–30 year) forecasts of expected changes in aboveground biomass (AGB). In this study, we forecast AGB changes across the North American boreal forest using machine learning, repeat measurements from 25,000 forest inventory sites, and gridded geospatial datasets. We find that AGB change can be predicted up to 30 years into the future, and that training on sites across the entire domain allows accurate predictions even in regions with only a small amount of existing field data. While predicting AGB loss is less skillful than gains, using a multi‐model ensemble can improve the accuracy in detecting change direction to >90% for observed increases, and up to 70% for observed losses. Higher stem density, winter temperatures, and the presence of temperate tree species in forest plots were positively associated with AGB change, whereas greater initial biomass, continentality (difference between mean summer and winter temperatures), prevalence of black spruce (Picea mariana), summer precipitation, and early warning metrics from long‐term remote sensing time series were negatively associated with AGB change. Across the domain, we predict nondisturbance‐induced declines in AGB at 23% of sites by 2030. The approach developed here can be used to estimate near‐future forest biomass in boreal North America and inform relevant management decisions. Our study also highlights the power of machine learning multi‐model ensembles when trained on a large volume of forest inventory plots, which could be applied to other regions with adequate plot density and spatial coverage. 

Abstract Rapid climate warming has contributed to significant changes in Arctic and boreal vegetation over the past half century. Changes in vegetation can impact wildlife by altering habitat and forage availability, which can affect behavior and range use. However, animals can also influence vegetation through foraging and trampling and therefore play an important role in determining ecosystem responses to climate change. As wildlife populations grow, density‐dependent processes can prompt range expansion or shifts. One mechanism for this is density‐dependent forage reduction, which can contribute to nutritional stress and population declines, and can also alter vegetation change trajectories. We assessed the range characteristics of a migratory caribou (Rangifer tarandus) herd in east‐central Alaska and west‐central Yukon Territory as it grew (1992–2017) then declined (2017–2020). Furthermore, we analyzed the correlation between caribou relative spatial density and vegetation change over this period using remotely sensed models of plant functional type cover. Over this period, caribou population density increased in all seasonal ranges. This was most acute in the calving range where density increased 8‐fold, from 1.5 to 12.0 animals km⁻². Concurrent with increasing density, we documented range shifts and expansion across summer, post‐calving and winter ranges. In particular, summer range size doubled (12,000 km²increase) and overlap with core range (areas with repeated year‐round use) was halved. Meanwhile, lichen cover, a key forage item, declined more in areas with high caribou density (2.4% absolute, 22% relative decline in cover) compared to areas where caribou were mostly absent (0.3% absolute, 1.9% relative decline). Conversely, deciduous shrub cover increased more in high caribou density areas. However, increases were dominated by less palatable shrubs whereas more palatable shrubs (i.e., willow [Salixspp.]) were stable or declined slightly. These changes in vegetation cover were small relative to uncertainty in the map products used to calculate change. Nonetheless, correlations between vegetation change and caribou range characteristics, along with concerning demographic trends reported over this same period, suggest changing forage conditions may have played a role in the herd's subsequent population decline. Our research highlights the potential of remotely sensed metrics of vegetation change for assessing the impacts of herbivory and trampling and stresses the importance of in situ data such as exclosures for validating such findings. 

Abstract Changes in vegetation distribution are underway in Arctic and boreal regions due to climate warming and associated fire disturbance. These changes have wide ranging downstream impacts—affecting wildlife habitat, nutrient cycling, climate feedbacks and fire regimes. It is thus critical to understand where these changes are occurring and what types of vegetation are affected, and to quantify the magnitude of the changes. In this study, we mapped live aboveground biomass for five common plant functional types (PFTs; deciduous shrubs, evergreen shrubs, forbs, graminoids and lichens) within Alaska and northwest Canada, every five years from 1985 to 2020. We employed a multi-scale approach, scaling from field harvest data and unmanned aerial vehicle-based biomass predictions to produce wall-to-wall maps based on climatological, topographic, phenological and Landsat spectral predictors. We found deciduous shrub and graminoid biomass were predicted best among PFTs. Our time-series analyses show increases in deciduous (37%) and evergreen shrub (7%) biomass, and decreases in graminoid (14%) and lichen (13%) biomass over a study area of approximately 500 000 km². Fire was an important driver of recent changes in the study area, with the largest changes in biomass associated with historic fire perimeters. Decreases in lichen and graminoid biomass often corresponded with increasing shrub biomass. These findings illustrate the driving trends in vegetation change within the Arctic/boreal region. Understanding these changes and the impacts they in turn will have on Arctic and boreal ecosystems will be critical to understanding the trajectory of climate change in the region. 

Abstract Widespread changes in the distribution and abundance of plant functional types (PFTs) are occurring in Arctic and boreal ecosystems due to the intensification of disturbances, such as fire, and climate-driven vegetation dynamics, such as tundra shrub expansion. To understand how these changes affect boreal and tundra ecosystems, we need to first quantify change for multiple PFTs across recent years. While landscape patches are generally composed of a mixture of PFTs, most previous moderate resolution (30 m) remote sensing analyses have mapped vegetation distribution and change within land cover categories that are based on the dominant PFT; or else the continuous distribution of one or a few PFTs, but for a single point in time. Here we map a 35 year time-series (1985–2020) of top cover (TC) for seven PFTs across a 1.77 × 10 6 km 2 study area in northern and central Alaska and northwestern Canada. We improve on previous methods of detecting vegetation change by modeling TC, a continuous measure of plant abundance. The PFTs collectively include all vascular plants within the study area as well as light macrolichens, a nonvascular class of high importance to caribou management. We identified net increases in deciduous shrubs (66 × 10 3 km 2 ), evergreen shrubs (20 × 10 3 km 2 ), broadleaf trees (17 × 10 3 km 2 ), and conifer trees (16 × 10 3 km 2 ), and net decreases in graminoids (−40 × 10 3 km 2 ) and light macrolichens (−13 × 10 3 km 2 ) over the full map area, with similar patterns across Arctic, oroarctic, and boreal bioclimatic zones. Model performance was assessed using spatially blocked, nested five-fold cross-validation with overall root mean square errors ranging from 8.3% to 19.0%. Most net change occurred as succession or plant expansion within areas undisturbed by recent fire, though PFT TC change also clearly resulted from fire disturbance. These maps have important applications for assessment of surface energy budgets, permafrost changes, nutrient cycling, and wildlife management and movement analysis. 

This dataset provides estimates of live, oven-dried aboveground biomass of all plants (tree, shrub, graminoid, forb, bryophyte) and all woody plants (tree, shrub) at 30-meter resolution across the Arctic tundra biome. Estimates of woody plant dominance are also provided as: (woody plant biomass / plant biomass) * 100. Plant biomass and woody plant biomass were estimated for each pixel (grams per square meter [g / m2]) using field harvest data for calibration/validation along with modeled seasonal surface reflectance data derived using Landsat satellite imagery and the Continuous Change Detection and Classification algorithm, and other supplementary predictors related to topography, region (e.g. bioclimate zone, ecosystem type), land cover, and derivative spectral products. Modeling was performed in a two-stage process using random forest models. First, biomass presence/absence was predicted using probability forests. Then, biomass quantity was predicted using regression forests. The model outputs were combined to produce final biomass estimates. Pixel uncertainty was assessed using Monte Carlo iterations. Field and remote sensing data were permuted during each iteration and the median (50th percentile, p500) predictions for each pixel were considered best estimates. In addition, this dataset provides the lower (2.5th percentile, p025) and upper (97.5th percentile, p975) bounds of a 95% uncertainty interval. Estimates of woody plant dominance are not modeled directly, but rather derived from plant biomass and woody plant biomass best estimates. The Pan Arctic domain includes both the Polar Arctic, defined using bioclimate zone data from the Circumpolar Arctic Vegetation Mapping Project (CAVM; Walker et al., 2005), and the Oro Arctic (treeless alpine tundra at high latitudes outside the Polar Arctic), defined using tundra ecoregions from the RESOLVE ecoregions dataset (Dinerstein et al., 2017) and treeline data from CAVM (CAVM Team, 2003). The mapped products focus on Arctic tundra vegetation biomass, but the coarse delineation of this biome meant some forested areas were included within the study domain. Therefore, this dataset also provides a tree mask product that can be used to mask out areas with canopy height ≥ 5 meters. This mask helps reduce, but does not eliminate entirely, areas of dense tree cover within the domain. Users should be cautious of predictions in forested areas as the models used to predict biomass were not well constrained in these areas. This dataset includes 132 files: 128 cloud-optimized GeoTIFFs, 2 tables in comma-separated values (CSV) format, 1 vector polygon in Shapefile format, and one figure in JPEG format. Raster data is provided in the WGS 84 / North Pole LAEA Bering Sea projection (EPSG:3571) at 30 meter (m) resolution. Raster data are tiled with letters representing rows and numbers representing columns, but note that some tiles do not contain unmasked pixels. We included all tiles nonetheless to maintain consistency. Tiling information can be found in the ‘metadata’ directory as a figure (JPEG) or shapefile.