Within the carbon cycle feedback, mechanisms that amplify or dampen the exchange of carbon dioxide between the different reservoirs to enhance concentrations of carbon dioxide and increase temperature from anthropogenic perturbations, play a crucial role. Quite a lot of these feedbacks are known, but most of them only potentially. This article evaluates the role of a number of these feedback processes within the carbon cycle. In order to assess their impact, some terrestrial feedbacks have been built into a coupled carbon cycle and climate model, as part of the integrated climate assessment model IMAGE.
A number of simulation experiments have been performed with this coupled carbon cycle/climate model to compare historical atmospheric concentration values of carbon dioxide with simulated values. Also global biospheric and oceanic carbon fluxes were validated against other modelling estimates. Based on the assumptions of the IPCC's 1990 Business-as-Usual (BaU-1990) scenario, future projections of the carbon dioxide concentration have been made. A key principle in this is that we have used the modelled feedbacks in order to balance the past and present carbon budget. For atmospheric carbon dioxide, this results in substantially lower projections than the IPCC-estimates: the difference in 2100 is approximately 16% from the 1990 level. Furthermore, the IPCC's 'best guess' or 'central estimate' value of the CO2 concentration in 2100 falls outside the uncertainty range estimated with our balanced modelling approach. Sensitivity experiments with the model have been performed to quantify to what extent the terrestrial feedback processes and oceanic fluxes influence the global carbon balance in the model. It is shown that a historical and present carbon balance can be obtained in many different ways, resulting in different biospheric fluxes and thus in considerably different atmospheric CO2 projections.