Pan-tropical analysis of deforestation drivers using global land cover products and the role of deforestation drivers in global gross carbon emissions for 1990-2000-2005 time period

Organised by Laboratory of Geo-information Science and Remote Sensing

Thu 12 March 2015 12:45 to 13:15

Venue Atlas, gebouwnummer 104
Room 1

by Simon Besnard (France)


Tropical deforestation is considered the second largest source of anthropogenic greenhouse gas emissions. Consequently, there is a need to identify land use change and forestry activities - in particular those that are linked to the drivers of deforestation and forest degradation - and to assess their potential contribution to the mitigation of climate change. In this study, we aimed to (i) assess the suitability of global land cover maps for deforestation drivers’ analysis and (ii) to analyse gross carbon emissions per driver using sample-based driver data and global forest biomass maps in the pan-tropical region for the 1990-2005 time period. First, our study assessed three global land cover maps for deforestation drivers’ classification in Africa and South America: Modis500, CCI-LC and GLC-SHARE land cover maps. We showed that there are opportunities to use global land cover maps for deforestation drivers’ analysis, in particular with the CCI-LC product in the two continents. The estimated overall accuracies of the CCI-LC products were 60% and 90% in Africa and South America, respectively. User and producer’s accuracies for each driver class were also reported. We concluded that the CCI-LC product could be used for the agriculture driver classification in both continents although this product will overmap agriculture class in Africa- i.e. low user accuracy. Second, we estimated carbon losses per driver in South America for the 1990-2005 time period based on the 2010 global remote sensing survey of the FRA2010 and the deforestation drivers classification dataset. Forest cover changes per driver were combined with Yong et al. (2014) pan-tropical biomass map to estimate carbon losses. The method was validated with high-resolution carbon maps for both Colombia and Peru. We showed that there were gross carbon losses from deforestation of 195±4.2 MtC.yr-1 and 246±4.9 MtC.yr-1 for the 1990s and the 2000s, respectively. Agriculture account for circa 90% of the total carbon emissions for the two periods. Our study also demonstrated that circa 80% of the carbon losses from deforested occurred in the tropical rainforest ecosystem. Our estimates of carbon losses from deforestation are significantly lower compared to earlier published estimates.

Keywords: Climate change; tropical deforestation; deforestation drivers’ classification; global land cover maps; carbon emissions from deforestation; pantropical biomass maps.