FIGURE 3.6 Schematic showing the predicted long-term response over 100,000 years of atmospheric CO2, including ocean temperature feedbacks, to a range of possible fossil fuel emissions totals. The 100,000-year simulations include silicate weathering (solid lines) and the 35,000-year simulations include seafloor CaCO3 dissolution (dashed lines). These models highlight how the timescale of carbon uptake becomes extended as the event unfolds. Fast processes such as ocean uptake and biomass growth, with high transfer rates but limited capacity, lose their potency, while slower processes, such as seafloor carbonate dissolution and rock weathering, come to dominate.
SOURCE: Modified from Archer (2005).
only if they can be evaluated against observation. The historical record, and even the expanse of the Quaternary climate record, contains nothing comparable.
Observations and modeling of the past carbon cycle perturbations provide a basis for projecting future conditions under the full range of fossil fuel burning scenarios, including the most pessimistic “business-as-usual” eventuality. Along this trajectory, atmospheric CO2 levels will rise as long as fossil fuel burning continues (with ultimate input of ~5,000 GtC), rising to levels perhaps as high as 1,600-2,000 parts per million (i.e., five to seven times the preindustrial level) (Figure 3.6). The geological record of past hyperthermal events, including the PETM, suggests that severe global warming under such magnitudes of carbon emissions will persist for 20,000 to 40,000 years. Carbon cycle models indicate that even after 100,000 years, the anthropogenic perturbation to the carbon cycle will still be important, especially if the total amount of carbon emitted is large. Consequently, Milankovitch forcings that have so dominated the pacing and extent of climate variations, and especially ice sheets, over the last 2 million years will—as they were prior to the onset of the current