|Space-based Information:||This data set represents a prototype high resolution land cover map at 20m over Africa based on 1 year of Sentinel-2A observations from December 2015 to December 2016.
, The legend of the S2 prototype LC 20m map of Africa 2016 was built after reviewing various existing typologies (e.g. LCCS, LCML…), global (e.g. GLC-share, GlobeLand30) and national experiences (Africover, SERVIR-RMCD). The legend includes 10 generic classes that appropriately describe the land surface at 20m: "trees cover areas", "shrubs cover areas", "grassland", "cropland", "vegetation aquatic or regularly flooded", "lichen and mosses / sparse vegetation", "bare areas", "built up areas", "snow and/or ice" and "open water". , Two classification algorithms, the Random Forest (RF) and Machine Learning (ML), were chosen to transform the cloud-free reflectance composites generated by the pre-processing module into a land cover map. The two maps resulting from both approaches are then combined either to select the best representation of a land cover class which will be part of the final S2 prototype LC 20m map of Africa 2016 or, in case of unreliable LC class delineation, the reference layer is used to consolidate the land cover classification.