Land Cover Classification from Sentinel-1 and Sentinel-2, Western Siberia, Northern Part

Bartsch, Annett; Widhalm, Barbara

Sentinel-2 bands 2, 3, 4, 8, 11, 12 and Sentinel-1 minimum backscatter and slope (angular signature) have been combined in an unsupervised classification. For the classification clouds and water have been masked out and water was additionally masked by using the water class of GLC30. Furthermore areas with NDVI < 0.4 were excluded. The classes have been labelled based on two different classification schemes. Labels are inconsistent for some classes. The current version provides in such cases naming from both schemes separated by ‘/’ or ‘;’ in the meta data. The water layer of the ‘fmask’ algorithm was added to the final product and areas with NDVI < 0.4 were identified as ‘dry cryptogamic-crust/ naturally bare land’.

To request access to data please contact GlobPermafrost.

Interested parties will be granted the user password.

Data and Resources

Additional Info

Field Value
Project(s) ESA GlobPermafrost
Institute ZAMG Zentralanstalt für Meteorologie und Geodynamik
Publication Date
Version 2.0
Product group Land Cover Prototype LCP
Product Landcover Classes
Sensor Sentinel-1, Sentinel-2
  • Landcover Classes
Variables [Units]
  • 0 - water
  • 1 - dry cryptogamic-crust; naturally bare land
  • 2 - moist graminid, prostrate to erect dwarf shrub tundra; shrub moss tundra
  • 3 - moist graminid, prostrate dwarf shrub tundra
  • 4 - dry postrate dwarf-shrub lichen tundra
  • 5 - moist graminid, prostrate dwarf-shrub tundra/ willow complex, shrubby peatland, young stand
  • 6 - moist to wet shrub tundra / moist graminid, prostrate dwarf-shrub tundra/ wet graminoid, moss tundra, human impact
  • 7 - wet graminoid, moss tundra
  • 8 - Meadows, Grass and herb-dominated non-peatlands/Willow-complex
  • 9 - spruce to mixed forest
Region Western Siberia
Spatial Reference EPSG:32652 WGS 84 / UTM Zone 42, WGS84
Spatial Resolution 20 m
Spatial Coverage
Temporal Coverage
Temporal Resolution
Format Geotiff
Intra-annual observation period July - August
end_time 2017-08
start_time 2016-07
units ["Classes"]

Dataset extent

Map tiles by Stamen Design, under CC BY 3.0. Data by OpenStreetMap, under CC BY SA.