A Methodology for Quantifying Uncertainty in Climate Projections

Possible climate change caused by an increase in greenhouse gas concentrations, despite having been a subject of intensive study in recent years, is still very uncertain. Uncertainties in projections of different climate variables are usually described only by the ranges of possible values. For assessing the possible impact of climate change, it would be more useful to have probability distributions for these variables. Obtaining such distributions is usually very computationally expensive and requires knowledge of probability distributions for characteristics of the climate system that affect climate projections. A few studies of this kind have been carried out with energy balance/upwelling diffusion models. Here we demonstrate a methodology for performing a similar study with a 2 dimensional (zonally averaged) climate model that reproduces the behavior of coupled atmosphere/ocean general circulation models more realistically than energy balance models. This methodology involves application of the Deterministic Equivalent Modeling Method to derive functional approximations of the model’s probabilistic response. Monte Carlo analysis is then performed on the approximations. An application of the methodology is demonstrated by deriving the uncertainty in surface air temperature change and sea level rise due to thermal expansion of the ocean that result from uncertainties in climate sensitivity and the rate of heat uptake by the deep ocean for a prescribed increase in atmospheric CO2 concentration. We also demonstrate propagation of correlated uncertainties through different models, by presenting results that include uncertainty in projected carbon emissions.