An interesting component of the most recent coupled model inter-comparison project (CMIP5) is the inclusion of so-called “idealized” climate change experiments. These use simplified versions of the fully-coupled climate models used to perform the 21st century projections and other transient climate change experiments, such as either steadily or instantaneously increasing the atmospheric CO2 concentration. In the first of these, known as “amip” experiments, there is no coupling between the atmosphere and the oceans: in the control experiment the sea-surface temperatures (SSTs) are prescribed from observations over the 1979-2008 period and the atmospheric model responds to these. The climate change experiment then repeats this scenario except that the SSTs are increased: in one case by uniformally adding 4ºC to the SSTs everywhere, in another by imposing a geographical pattern of SST change. The second experiment is even more idealized. These are “aquaplanet” (water world ) simulations: there is no land, the prescribed SSTs are zonally symmetric and the solar insolation is for a fixed season. The climate change experiment with these models then consists of simply increasing the SSTs by 4ºC everywhere.
In our new paper, just published in Geophysical Research Letters, we calculate the global-mean radiative feedbacks in these idealized experiments and compare them to the values obtained from the fully-coupled models in response to increased CO2. We find that in all but two cases − two of the aquaplanet models − the cloud feedbacks match up almost perfectly (right hand figure). This is an interesting result because it suggests that factors such as the spatial pattern of the SST increase, the increased surface warming over land and the coupling between the atmosphere and ocean may not be especially relevant to determining the global-mean cloud feedbacks in models. It also implies that the simplified experiments should provide a useful framework for understanding the processes governing the cloud feedbacks. Note that the total radiative feedback is not so well characterized by the idealized experiments (left hand figure): this is primarily because they do not include the positive feedback due to reduced sea ice in response to warming. The paper also examines the relationship between the feedbacks and the CO2 radiative forcing. In particular we find an anti-correlation between the rapid cloud adjustments in response to increased CO2 and the cloud feedbacks. This strengthens as the experiments get progressively simpler, so that it is strongest in the aquaplanet model ensemble.
The abstract and citation are below. The paper can be accessed here.
Analysis of the available Coupled Model Intercomparison Project Phase 5 models suggests that sea surface temperature-forced, atmosphere-only global warming experiments (“amip4K,” “amipFuture,” and “aqua4K”) are a good guide to the global cloud feedbacks determined from coupled atmosphere-ocean CO2-forced simulations, including the intermodel spread. Differences in the total climate feedback parameter between the experiments arise primarily from differences in the clear-sky feedbacks which can largely be anticipated from the nature of the experimental design. The effective CO2 radiative forcing is anticorrelated with the total feedback in the coupled simulations. This anticorrelation strengthens as the experimental design becomes simpler, the number of potential degrees of freedom of the system’s response reduces, and the relevant physical processes can be identified. In the aquaplanet simulations the anticorrelation is primarily driven by opposing changes in the rapid cloud adjustment to CO2 and the net cloud response to increased surface warming. Establishing a physical explanation for this behavior is important future work.
Ringer, M. A., T. Andrews, and M. J. Webb (2014), Global-mean radiative feedbacks and forcing in atmosphere-only and coupled atmosphere-ocean climate change experiments, Geophys. Res. Lett., 41, doi:10.1002/2014GL060347.
Mark Ringer 