scieee AI-readable full text Open interactive document viewer

Hybrid Land Cover Product: A hybrid circumpolar 1 km land cover product

Briones, Valeria; Genet, Helene; Jafarov, Elchin E.; Rogers, Brendan; Watts, Jennifer D.; Virkkala, Anna-Maria; Bartsch, Annett; Maglio, Benjamin C.; Natali, Susan M.; Rady, Joshua

Abstract

This 1km hybrid land cover map of the Arctic-Boreal region was created by integrating global and regional datasets using a machine learning-based fusion approach. It includes 35 classes such as forest types, shrublands, and wetlands, and captures the fine-scale spatial heterogeneity essential for understanding permafrost dynamics. The product improves upon existing global maps by offering greater thematic detail and consistency across the circumpolar Arctic, supporting ecological research and climate modeling. **This dataset supports the manuscript "Fusing Regional and Global Land Cover Maps: An Application for Ecosystem Modeling Across High Latitudes" which is currently under review.A DOI for the paper will be added here upon publication.

Full text

Fusing Regional and Global Land Cover Maps: An Application for Ecosystem Modeling Across High Latitudes Valeria Briones1, Hélène Genet2, Elchin E. Jafarov1, Brendan M. Rogers1, Jennifer D.Watts1, Anna-Maria Virkkala1 Annett Bartsch3, Benjamin C. Maglio2, Joshua Rady1, Susan M. Natali1 1 Woodwell Climate Research Center, Falmouth, MA, 02540, USA 2Institute of Arctic Biology, University of Alaska Fairbanks, AK 99775 Fairbanks, USA 3 b.geos, Industriestrasse 1, 2100 Korneuburg, Austria E-mail: [email protected]g Land cover is a critical dataset for studying permafrost dynamics, carbon cycling, ecosystem resilience, and climate feedbacks in high-latitude environments. However, global land cover products often fail to capture the spatial heterogeneity and ecological detail required for Arctic-Boreal applications. This circumpolar 1-km land cover map integrates global (ESA-CCI-LC) and regional datasets using a machine learning-based fusion approach to generate 35 detailed land cover classes, including forest, shrubland, and wetland types. The map ensures thematic richness and consistency across the Arctic-Boreal zone, supporting improved climate modeling and ecological analysis. Further methodological details can be found in the associated publication. Figure 1. The overview of the hybrid land cover product Class Class Name 6 Poplar Forest 7 Oak Forest 10 Other Pine Forest 11 Fir Forest 12 Aspen Forest 13 Hemlock Forest 14 Spruce Forest 15 White Spruce Forest 16 Black Spruce Forest 17 Mixed Forest 18 Larch (Tamarack) Forest 19 Birch Forest 20 Scots Pine Forest 21 Jack Pine Forest 22 Cedar Forest 23 Maple Forest 24 Siberian Pine Forest 25 Linden Forest 26 Cedar Elfin Wood Forest 30 Alpine Shrubland 31 Riparian Shrubland 32 Other Shrubland 33 Herbaceous 35 Sparsely Vegetated 50 Bog 51 Fen 52 Marsh 60 Erect-shrub tundra 61 Prostrate-shrub tundra 62 Shrub tundra 63 Graminoid tundra 64 Wet-sedge tundra 65 Barren tundra 70 Developed 96 Bare, Rock 97 Other 98 Water 99 Snow/Ice Projection information: ----------------------- Spatial reference: WGS 1984 NSIDC EASE-Grid 2.0 North Linear unit: Meter Angular unit: Degree False easting: 0 False northing: 0 Central meridian: 0 Latitude_of_origin: 90