Key points are not available for this paper at this time.
Abstract This research refers to an object‐based automatic method combined with a national expert validation to produce regional and national forest cover change statistics over C ongo B asin. A total of 547 sampling sites systematically distributed over the whole humid forest domain are required to cover the six C entral A frican countries containing tropical moist forest. High resolution imagery is used to accurately estimate not only deforestation and reforestation but also degradation and regeneration. The overall method consists of four steps: (i) image automatic preprocessing and preinterpretation, (ii) interpretation by national expert, (iii) statistic computation and (iv) accuracy assessment. The annual rate of net deforestation in C ongo B asin is estimated to 0.09% between 1990 and 2000 and of net degradation to 0.05%. Between 2000 and 2005, this unique exercise estimates annual net deforestation to 0.17% and annual net degradation to 0.09%. An accuracy assessment reveals that 92.7% of tree cover ( TC ) classes agree with independent expert interpretation. In the discussion, we underline the direct causes and the drivers of deforestation. Population density, small‐scale agriculture, fuelwood collection and forest's accessibility are closely linked to deforestation, whereas timber extraction has no major impact on the reduction in the canopy cover. The analysis also shows the efficiency of protected areas to reduce deforestation. These results are expected to contribute to the discussion on the reduction in CO 2 emissions from deforestation and forest degradation ( REDD +) and serve as reference for the period.
Ernst et al. (Tue,) studied this question.