A vegetation map of Central Africa derived from satellite imagery

Citation
P. Mayaux et al., A vegetation map of Central Africa derived from satellite imagery, J BIOGEOGR, 26(2), 1999, pp. 353-366
Citations number
35
Categorie Soggetti
Environment/Ecology
Journal title
JOURNAL OF BIOGEOGRAPHY
ISSN journal
03050270 → ACNP
Volume
26
Issue
2
Year of publication
1999
Pages
353 - 366
Database
ISI
SICI code
0305-0270(199903)26:2<353:AVMOCA>2.0.ZU;2-4
Abstract
Aim This paper presents the Joint Research Centre's TREES Project satellite derived Vegetation Map of Central Africa, at 1:5,000,000 scale, with a det ailed description of the vegetation classes and their distribution. The inf ormation content of the map is compared with other conventional and satelli te derived maps of the region for validation and evaluation purposes. Location The map focuses on the Guineo-Congolian ecological domain and cove rs the following countries: Cameroon, Central African Republic, Republic of Congo, Equatorial Guinea, Gabor 1 and the Democratic Republic of Congo. Methods Using coarse resolution satellite imagery a map of vegetation cover has been produced based upon the spectral response of the vegetation cover . Digital image processing and geographical information systems techniques were employed, together with local knowledge, high resolution imagery and e xpert consultation, to compile a cartographic map product. Results The TREES Vegetation Map of Central Africa has been shown to be str ongly correlated with the FAO Forest Resources Assessment for 1990. Compari son with other map sources indicates that the map contains greater spatial detail and is more consistent than conventionally compiled maps. The conven tional maps however, contain more thematic information content relating to vegetation type. Main conclusions The map improves our stale of knowledge of the vegetation cover of Central Africa and presents the most consistent and spatially deta iled view yet published at this scale. Thematic information content on fore st type is limited but should be improved in the near future with the inclu sion of data from new satellite sensors. This first version of the map and future planned updates should provide an important input for regional strat ification and planning purposes for forest resources, biodiversity and clim ate studies.