Méthodologie De Production Des Données Cartographiques Sur Le Couvert Forestier En République Du Congo, De 2000 À 2012

Abstract

A forest-based mapping data production study was conducted in the Republic of Congo as part of the Reducing Emissions from Deforestation and Forest Degradation (REDD) process including sustainable forest management, forest conservation, and forest degradation. biodiversity and increasing carbon stocks (REDD +). The purpose of this study was to produce a forest cover map and cover changes for the period 2000 to 2012, which meets the criteria of the national definition of forest adopted by the Republic of Congo and the evaluation of forest areas. and losses. The map produced was based on the combination of forest map parameters developed prior to the adoption of the forest definition criteria. The 2000 and 2010 Landsat image segmentation was used to create polygons that respect the definition and in which the values of the underlying maps were injected, through a decision tree transcribed in the R language and implemented in conjunction with the tools Open Foris Tools Kit. The results indicate an overall accuracy of 90% of the map without stratification, with greater accuracy for the forest (86%) than for the losses (73%). The forest occupies 23,517,000 hectares, representing 69% of the national territory. The forest area lost during this period was estimated at about 145,356 hectares, which represents an annual average of 12,113 hectares (0.052%). These results lead to the conclusion that the Republic of Congo is one of the countries with high forest cover and low deforestation.