Predicting the Long Term Future

Linear and nonlinear analyses are performed on a generic carbon cycle model, exploiting available empirical information and avoiding detailed physics as far as possible. It is shown that the ability to reproduce atmospheric CO2 historical data of past decades is neither necessary nor sufficient to establish credibility for the century-timescale predictions. This is because the relevant historical data is driven by a “rapid” exponential forcing term, while the future forcing term of interest is expected to be flat and slowly time varying. Simple analytical formulas relating long term atmospheric CO2 level and anthropogenic emission are derived, involving only three parameters. In the centurytimescale, there exists no known quick way to empirically determine or validate the values of these parameters. An estimate of the time to breach an user-specified threshold of atmospheric CO2 level in terms of these parameters is provided. ∗The author thanks Professors R. Socolow and F. Dryer for their encouragements and critiques during the course of the research.