Berthelot,M., P. Ciais, P. Friedlingstein, J-L. Dufresne, and P. Monfray, 2003:
How uncertainties in future climate change predictions translate into future terrestrial carbon fluxes
Global Change Biology (submitted)
We forced a global terrestrial carbon cycle model by climate fields of 14 Ocean-Atmosphere General Circulation Models (OAGCM) to simulate the response of terrestrial carbon pools and fluxes to climate change over the next century. These models participated in the second phase of the Coupled Model Intercomparison Project (CMIP2) where a 1% per year increase of atmospheric CO2 was prescribed. We obtain a reduction in net land uptake (NEP) because of climate change, from 1.4 to 5.7 GtC per year at the time of atmospheric CO2 doubling. Such reduction in terrestrial carbon sinks is largely dominated by the response of tropical ecosystems, where soil water stress occurs. The uncertainty in the simulated land carbon cycle response is the consequence of discrepancies in land temperature and precipitation changes as simulated by the OAGCM. We use a statistical approach, based on a factor called interexperiment agreement, to determine how coherent the responses of land carbon fluxes to differents scenarios of climate change are. At the continental scale, the interexperiment agreements of the climate impact on biospheric carbon fluxes and pools changes are good in the tropics, in the Mediterranean region and in high latitudes of Northern Hemisphere. These results are the consequence of a good inter-experiment agreement of soil water content change, which controls the carbon exchange between land biosphere and atmosphere in the tropics and in the Mediterranean region and of temperature change in the Northern Hemisphere. The remaining disagreement between the terrestrial responses to climate change is due to a combination of differences between the 14 OAGCMs climates and of internal OAGCM climate variability. We show that disagreement of simulated Net Primary Production (NPP) and carbon pools changes are mainly driven by intermodel differences whereas the NEP disagreement is however driven by internal variability. We then evaluate the carbon uptake uncertainties linked to assumptions on plant productivity sensitivity to atmospheric CO2 and on decomposition rate sensitivity to temperature. We show that this uncertainty is on the same order of magnitude than the uncertainty due to the simulated climate change. Finally, we find that the OAGCMs having the largest climate sensitivities to CO2 increase are also those with the largest soil drying in the tropics, and therefore those with the largest reduction of carbon uptake.