Climate sensitivity and the rate of heat uptake by the deep ocean are believed to be the most important characteristics defining the transient response of climate models to external forcings. Differences between results of climate change simulations produced by existing coupled AOGCMs are in large part due to differences in these two characteristics. The direct comparison of climate change projections from different AOGCMs, however, does not allow one to separate contributions of these two parameters.
A 2D climate model was developed in the framework of the MIT Joint Program on the Science and Policy of Global Change for studying uncertainties in climate change projections. This model consists of a two-dimensional (zonal averaged) atmospheric model developed from the GISS atmospheric GCM coupled to a diffusive ocean model. The model's sensitivity can be varied by changing the strength of the cloud feedback. As was shown in Sokolov and Stone (1998) the behavior of different coupled AOGCMs can be reproduced by simulations with the 2D climate model with an appropriate choice of the model's sensitivity and coefficient for heat diffusion in the deep ocean. It is important that we are able to reproduce responses of a given AOGCM to different forcings with a unique choice of these parameters.
There are two ways to estimate the values of these parameters that allow us to match the behavior of a given AOGCM. The first is to use the model's climate sensitivity obtained from equilibrium climate change simulations (usually a doubled CO2 simulation) with the atmospheric sub-model coupled to a mixed layer ocean model together with surface air temperature from a run with the coupled AOGCM. Second, the values of the parameters in question can be defined from data for the surface air temperature and the thermal expansion of ocean from a run with the coupled AOGCM. The data for thermal expansion are available for at least some AOGCMs.
As discussed in Sokolov and Stone (1998) and Webster and
Sokolov (2000), uncertainty in the rate of oceanic heat uptake
plays an important role in defining the overall uncertainty in
climate change projections. A methodology for studying
uncertainty in climate change projections as described in Webster
and Sokolov requires knowledge of the probability distributions for
different parameters including the rate of heat uptake by the deep
ocean. The distribution used in the above study was based on the
results of climate change simulations with a rather few coupled
AOGCMs. The proposed sub-project will allow us to construct
more reliable distributions for use in future uncertainty studies and
therefore obtain better estimates of the relative importance of
uncertainty in oceanic heat uptake.
Sokolov, A.P. and P.H. Stone, 1998: "A Flexible Climate Model for Use in Integrated Assessments", Climate Dynamics 14:291-303.
Webster M. and A. Sokolov, 2000: "A Methodology for Quantifying the Uncertainty in Climate Projections." Climatic Change (in press)