The precipitation analysis shows that the models share only the broadest aspects of the seasonal cycle in precipitation and the common vector loadings vary substantially between the models. The analysis of the ensemble integrations for the same variable indicates that the model variations are generally quite a bit larger than the variations amongst the members of the ensemble. The ensemble results give some confidence that the differences among the models are robust and that they would not change greatly for other realizations of the model integrations. In itself, the ensemble analysis displays the usefulness of the CPC technique in succinctly combining the output of any number of realizations from a given model. Their common character is a robust signal and is the answer often sought from ensemble experiments.
The CPC analysis of the five ensemble simulations shows that the interannual variations in the divergence are dominated by the ENSO events for the AMIP decade, 1979-1988 The analysis indicates that the ENSO global signal is robust across all the simulations such that one simulation is all that is necessary to characterize the global response. The intrinsic variability of the model begins to dominate the components higher than the first.
The difference between the 200 hPa divergence of four closely related AMIP models and the NCEP /NCAR reanalysis was analyzed using the CPC technique. The models all share the ECMWF dynamical core. When compared to the analogous analysis of the ensembles, it is seen that the models form two distinct pairs although some distinctive characteristics are retained. The CPC is shown to have value in documenting the impact of modifying parameterizations on the simulations.
The CPC analyses tend to support the observation that the models often have more in common with each other than with the observations. The CPC approach has utility in answering many useful questions posed in the arena of model comparison when used in conjunction with other techniques. (pdf file)