Search for: All records

Abstract Ecosystems at high latitudes are changing rapidly in response to climate change. To understand changes in carbon fluxes across seasonal to multi‐decadal timescales, long‐term in situ measurements from eddy covariance networks are needed. However, there are large spatiotemporal gaps in the high‐latitude eddy covariance network. Here we used the relative extrapolation error index in machine learning‐based upscaled gross primary production as a measure of network representativeness and as the basis for a network optimization. We show that the relative extrapolation error index has steadily decreased from 2001 to 2020, suggesting diminishing upscaling errors. In experiments where we limit site activity by either setting a maximum duration or by ending measurements at a fixed time those errors increase significantly, in some cases setting the network status back more than a decade. Our experiments also show that with equal site activity across different theoretical network setups, a more spread out design with shorter‐term measurements functions better in terms of larger‐scale representativeness than a network with fewer long‐term towers. We developed a method to select optimized site additions for a network extension, which blends an objective modeling approach with expert knowledge. This method greatly outperforms an unguided network extension and can compensate for suboptimal human choices. For the Canadian Arctic we show several optimization scenarios and find that especially the Canadian high Arctic and north east tundra benefit greatly from addition sites. Overall, it is important to keep sites active and where possible make the extra investment to survey new strategic locations. 

Abstract Modeling Arctic-Boreal vegetation is a challenging but important task, since this highly dynamic ecosystem is undergoing rapid and substantial environmental change. In this work, we synthesized information on 18 dynamic vegetation models (DVMs) that can be used to project vegetation structure, composition, and function in North American Arctic-Boreal ecosystems. We reviewed the ecosystem properties and scaling assumptions these models make, reviewed their applications from the scholarly literature, and conducted a survey of expert opinion to determine which processes are important but lacking in DVMs. We then grouped the models into four categories (specific intention models, forest species models, cohort models, and carbon tracking models) using cluster analysis to highlight similarities among the models. Our application review identified 48 papers that addressed vegetation dynamics either directly (22) or indirectly (26). The expert survey results indicated a large desire for increased representation of active layer depth and permafrost in future model development. Ultimately, this paper serves as a summary of DVM development and application in Arctic-Boreal environments and can be used as a guide for potential model users, thereby prioritizing options for model development. 

Boreal and Arctic regions have warmed up to four times quicker than the rest of the planet since the 1970s. As a result, boreal and tundra ecosystems are experiencing more frequent and higher intensity extreme weather events and disturbances, such as wildfires. Yet limitations in ground and satellite data across the Arctic and boreal regions have challenged efforts to track these disturbances at regional scales. In order to effectively monitor the progression and extent of wildfires in the Arctic-boreal zone, it is essential to determine whether burned area (BA) products are accurate representations of BA. Here, we use 12 different datasets together with MODIS active fire data to determine the total yearly BA and seasonal patterns of fires in Arctic-boreal North America and Russia for the years 2001–2020. We found relatively little variability between the datasets in North America, both in terms of total BA and seasonality, with an average BA of 2.55 ± 1.24 (standard deviation) Mha/year for our analysis period, the majority (ca. 41%) of which occurs in July. In contrast, in Russia, there are large disparities between the products—GFED5 produces over four times more BA than GFED4s in southern Siberia. These disparities occur due to the different methodologies used; dNBR (differenced Normalized Burn Ratio) of short-term composites from Landsat images used alongside hotspot data was the most consistently successful in representing BA. We stress caution using GABAM in these regions, especially for the years 2001–2013, as Landsat-7 ETM+ scan lines are mistaken as burnt patches, increasing errors of commission. On the other hand, we highlight using regional products where possible, such as ABoVE-FED or ABBA in North America, and the Talucci et al. fire perimeter product in Russia, due to their detection of smaller fires which are often missed by global products. 

Abstract. The snow cover extent across the Northern Hemisphere has diminished, while the number of lightning ignitions and amount of burned area have increased over the last 5 decades with accelerated warming. However, the effects of earlier snow disappearance on fire are largely unknown. Here, we assessed the influence of snow disappearance timing on fire ignitions across 16 ecoregions of boreal North America. We found spatially divergent trends in earlier (later) snow disappearance, which led to an increasing (decreasing) number of ignitions for the northwestern (southeastern) ecoregions between 1980 and 2019. Similar northwest–southeast divergent trends were observed in the changing length of the snow-free season and correspondingly the fire season length. We observed increases (decreases) over northwestern (southeastern) boreal North America which coincided with a continental dipole in air temperature changes between 2001 and 2019. Earlier snow disappearance induced earlier ignitions of between 0.22 and 1.43 d earlier per day of earlier snow disappearance in all ecoregions between 2001 and 2019. Early-season ignitions (defined by the 20 % earliest fire ignitions per year) developed into significantly larger fires in 8 out of 16 ecoregions, being on average 77 % larger across the whole domain. Using a piecewise structural equation model, we found that earlier snow disappearance is a good direct proxy for earlier ignitions but may also result in a cascade of effects from earlier desiccation of fuels and favorable weather conditions that lead to earlier ignitions. This indicates that snow disappearance timing is an important trigger of land–atmosphere dynamics. Future warming and consequent changes in snow disappearance timing may contribute to further increases in western boreal fires, while it remains unclear how the number and timing of fire ignitions in eastern boreal North America may change with climate change. 

Abstract Accelerated warming of the Arctic can affect the global climate system by thawing permafrost and exposing organic carbon in soils to decompose and release greenhouse gases into the atmosphere. We used a process-based biosphere model (DVM-DOS-TEM) designed to simulate biophysical and biogeochemical interactions between the soil, vegetation, and atmosphere. We varied soil and environmental parameters to assess the impact on cryohydrological and biogeochemical outputs in the model. We analyzed the responses of ecosystem carbon balances to permafrost thaw by running site-level simulations at two long-term tundra ecological monitoring sites in Alaska: Eight Mile Lake (EML) and Imnavait Creek Watershed (IMN), which are characterized by similar tussock tundra vegetation but differing soil drainage conditions and climate. Model outputs showed agreement with field observations at both sites for soil physical properties and ecosystem CO₂fluxes. Model simulations of Net Ecosystem Exchange (NEE) showed an overestimation during the frozen season (higher CO₂emissions) at EML with a mean NEE of 26.98 ± 4.83 gC/m²/month compared to observational mean of 22.01 ± 5.67 gC/m²/month, and during the fall months at IMN, with a modeled mean of 19.21 ± 7.49 gC/m²/month compared to observation mean of 11.9 ± 4.45 gC/m²/month. Our results underscore the importance of representing the impact of soil drainage conditions on the thawing of permafrost soils, particularly poorly drained soils, which will drive the magnitude of carbon released at sites across the high-latitude tundra. These findings can help improve predictions of net carbon releases from thawing permafrost, ultimately contributing to a better understanding of the impact of Arctic warming on the global climate system. 

Abstract Deciduous tree cover is expected to increase in North American boreal forests with climate warming and wildfire. This shift in composition has the potential to generate biophysical cooling via increased land surface albedo. Here we use Landsat-derived maps of continuous tree canopy cover and deciduous fractional composition to assess albedo change over recent decades. We find, on average, a small net decrease in deciduous fraction from 2000 to 2015 across boreal North America and from 1992 to 2015 across Canada, despite extensive fire disturbance that locally increased deciduous vegetation. We further find near-neutral net biophysical change in radiative forcing associated with albedo when aggregated across the domain. Thus, while there have been widespread changes in forest composition over the past several decades, the net changes in composition and associated post-fire radiative forcing have not induced systematic negative feedbacks to climate warming over the spatial and temporal scope of our study. 

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