Uncertainty Associated with Harmonization of Global Land-Use Scenarios for the 5th IPCC Assessment

Climate models (CMs) being developed for the fifth Assessment Report (AR5) of the Intergovernmental Panel on Climate Change (IPCC) will address the combined effects of human activities, including land-use, on the coupled carbon-climate system. Determining the sources, and potential magnitude, of the uncertainty associated with the land-use data that these CMs will use provides several new challenges. To address these challenges and more, we have prepared a harmonized set of land-use data for each of the Representative Concentration Pathways being modeled by the four Integrated Assessment Models for IPCC AR5. Our datasets are prepared using our Global Land-use Model (GLM) that includes several model variables that govern model decision-making; the choice of these variables can alter the resulting land-use datasets that GLM produces. As part of this effort, we have recently undertaken a study of the sensitivity and uncertainty surrounding these choices by computing a harmonized land-use dataset using every possible combination of model variables - over 1600 combinations in total. For each of these harmonized land-use datasets, GLM ensures a smooth and consistent transition from the historical land-use reconstructions to the future land-use projections, grids (or re-grids) the data when necessary, spatially allocates national/regional wood harvest statistics, and computes all the resulting land-use states, and transitions between land-use states (which determine carbon fluxes), annually from 1500 to 2100 at half-degree (fractional) spatial resolution. Our uncertainty analysis quantifies the potential range of several key output metrics (resulting secondary land area, age of secondary lands, cumulative losses of biomass from the terrestrial biosphere, and total net change in terrestrial biomass) and also indicates which variables have the greatest potential impact on these metrics. The simulation start date, the priority of primary vs. secondary land for clearing, the inclusion of wood harvesting, and the inclusion of shifting cultivation are all shown to play a significant role in our analysis and therefore represent the most important processes and factors to consider when including land-use data in CMs.