Sensitivity of land use change emission estimates to historical land use and land cover mapping

International audienceThe carbon emissions from land use and land cover change (E LUC) are an important anthropogenic component of the global carbon budget. Yet these emissions have a large uncertainty. Uncertainty in historical land use and land cover change (LULCC) maps and their implementation in global vegetation models is one of the key sources of the spread of E LUC calculated by global vegetation models. In this study, we used the Organizing Carbon and Hydrology in Dynamic Ecosystems terrestrial biosphere model to investigate how the different transition rules to define the priority of conversion from natural vegetation to agricultural land affect the historical reconstruction of plant functional types (PFTs) and E LUC. First, we reconstructed 10 sets of historical PFT maps using different transition rules and two methods. Then, we calculated E LUC from these 10 different historical PFT maps and an additional published PFT reconstruction, using the difference between two sets of simulations (with and without LULCC). The total area of forest loss is highly correlated with the total simulated E LUC (R 2 = 0.83, P < 0.001) across the reconstructed PFT maps, which indicates that the choice of transition rules is a critical (and often overlooked) decision affecting the simulated E LUC. In addition to the choice of a transition rule, the initial land cover map and the reconstruction method for the reconstruction of historical PFT maps have an important impact on the resultant estimates of E LUC