Circumarctic landcover diversity considering wetness gradients

• Main Authors: Bartsch, Annett, Efimova, Aleksandra, Widhalm, Barbara, Muri, Xaver, von Baeckmann, Clemens, Bergstedt, Helena, Ermokhina, Ksenia, Hugelius, Gustaf, Heim, Birgit, Leibmann, Marina
• Format: Article in Journal/Newspaper
• Language: English
• Published: Copernicus Publications 2023
• Subjects: article; Verlagsveröffentlichung; Arctic; Climate change; permafrost; Tundra
• Online Access: https://doi.org/10.5194/egusphere-2023-2295; https://noa.gwlb.de/receive/cop_mods_00069886; https://noa.gwlb.de/servlets/MCRFileNodeServlet/cop_derivate_00068254/egusphere-2023-2295.pdf; https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2295/egusphere-2023-2295.pdf

Landcover heterogeneity information considering soil wetness across the entire Arctic tundra is of interest for a wide range of applications targeting climate change impacts and ecological research questions. Patterns potentially link to permafrost degradation and affect carbon fluxes. First a landcover unit retrieval scheme which provides unprecedented detail by fusion of satellite data using Sentinel-1 (synthetic aperture radar) and Sentinel-2 (multispectral) has been adapted. Patterns of lakes, wetlands, general soil moisture conditions and vegetation physiognomy are represented at 10 m nominal resolution. Units with similar patterns are identified with a k-means approach and documented through statistics derived from comprehensive in situ records for soils and vegetation (more than 3500 samples). The result goes beyond the capability of existing landcover maps which have deficiencies in spatial resolution, thematic content and accuracy. Wetness gradients have been eventually assessed and measures for landscape heterogeneity were derived north of the treeline. About 40 % of the area north of the treeline falls into three units of dry types with limited shrub growth. Wetter regions have higher landcover diversity than drier regions. 45 % of the Arctic landscape is highly heterogeneous with respect to wetness considering 1kmx1km units (representative scale of frequently used regional landcover and permafrost modelling products). Wetland areas cover on average 9 % and moist tundra types 32 %, what is potentially of relevance for methane flux upscaling.