Pierre Friedlingstein, IPSL/LSCE, France + contributions from L. Bopp, P. Cadule, C. Jones, S. Piao, and J. Randerson Process oriented evaluation of coupled climate-carbon cycle models 2 . 6 – 4 . 1 ° C 2 . 4 – 5 . 6 ° C 7 3 0 – 1 0 0 0 p p m C 4 MIP in IPCC AR4 Climate feedback on Carbon Cycle Climate projections Friedlingstein., et al, 2006 Meehl ., et al, 2007 Atmospheric CO2 difference (ppm) Change in Land Uptake (GtC) Surface Temperature Change (K) C4MIP in CMIP5 Multiple constrains from observations Atmospheric CO 2 (trends and var.) Leaf Area Index (Greening) Ocean Carbon Inventory (trend) Air-Sea CO 2 exchange (trends and var.) C 4 MIP Umbrella • Benchmark against 20th century • Constrains on processes driving climate-carbon feedback • Reduced uncertainty on Climate-carbon cycle gain Compatible emissions Climate projections I-LAMB • Positive feedback (larger warming), but large uncertainty in simulated atmospheric CO 2 . • Essentially driven by uncertainty in carbon cycle response to climate change • At that time, no proper model evaluation C4MIP models AOGCMs C4MIP models