UCL-Forest Classification (DR Congo)

Based on: A. Verhegghen, P. Defourny, "A new 300 m vegetation map for Central Africa based on multi-sensor times series", Third Recent Advance in Quantitative Remote Sensing, J.A. Sobrino (Ed.), Publicaciones de la Universitat de Valencia,Valencia, Spain, 2010.608 Vegetation types of 8 countries in Central Africa have been mapped thanks to a semi-automatic processing method based on temporal and spectral information from 19 months of ENVISAT MERIS FRS observation and 8 years of SPOT VEGETATION time series. The approach is based on a previous 1-km mapping effort for the Democratic Republic of Congo and on the lessons learnt from the ESA-GlobCover project. A land cover map with 20 vegetation classes was produced in five major steps: data compositing, seasonal stratification of the study zone, unsupervised classifications, automatic labelling and manual editing. The floristic composition and physiognomy of each vegetation type are described using the Land Cover Classification System developed by the FAO. This mapping exercise will be a reference document to deliver area estimates of the different forest types in a consistent way for DRCongo, Gabon, Cameroon, Equatorial Guinea, Central African Republic, Congo, Rwanda and Burundi. For the display of the NFMS portal, the original data has been reprojected to "latlong" projection, cropped, and masked for the area of the DRC.