******************************EXCERPT*****************************
. Discussion
Three equally plausible models give very different expectations for future CO2 trajectories under business as usual assumptions. No inference is possible at this time as to which model is “right” because the three models are virtually identical in the CO2 data period (Fig. 2) and the understanding of the carbon cycle in this context is not precise enough. The factors governing CO2 in the atmosphere may or may not lend themselves to long-term predictability even if they were understood better. It is clear, however, that simply using an exponential model because it fits the data represents an incomplete analysis, as other models fit equally well. The IPCC “best estimate” of 836 ppmv in 2100, which is equivalent to extrapolation of the exponential model, is indeter- minate and could just as easily be 569.8 or 672.5 ppmv (or even 747.7 ppmv by Hofmann et al., 2009), as found using equally likely models that fit the same data. These much lower “best estimate” values affect the IPCC “high” estimate, which is derived from the base estimate exponential model by adding a growth term (based on higher economic growth rates and other factors). Because projections of future climate depend on future CO2 (and other greenhouse gas) levels, a future value below the IPCC low estimate would preclude the more extreme climate change forecasts made by the IPCC.
One more Time we see the IPCC at work ....