Influence of carbon mapping and land change modelling on the prediction of carbon emissions from deforestation

The implementation of an international programme for reducing carbon emissions from deforestation and degradation (REDD) can help to mitigate climate change and bring numerous benefits to environmental conservation. Information on land change modelling and carbon mapping can contribute to quantify future carbon emissions from deforestation. However limitations in data availability and technical capabilities may constitute an obstacle for countries interested in participating in the REDD programme. This paper evaluates the influence of quantity and allocation of mapped carbon stocks and expected deforestation on the prediction of carbon emissions from deforestation. The paper introduces the conceptual space where quantity and allocation are involved in predicting carbon emissions, and then uses the concepts to predict carbon emissions in the Brazilian Amazon, using previously published information about carbon mapping and deforestation modelling. Results showed that variation in quantity of carbon among carbon maps was the most influential component of uncertainty, followed by quantity of predicted deforestation. Spatial allocation of carbon within carbon maps was less influential than quantity of carbon in the maps. For most of the carbon maps, spatial allocation of deforestation had a minor but variable effect on the prediction of carbon emissions relative to the other components. The influence of spatial carbon allocation reaches its maximum when 50% of the initial forest area is deforested. The method can be applied to other case studies to evaluate the interacting effects of quantity and allocation of carbon with future deforestation on the prediction of carbon emissions from deforestation.