全球变化知识资源中心

A cloud feedback diagnostic package is implemented in the Australian Community Climate and Earth-System Simulator General Circulation Model, based on the methodology of cloud radiative kernels. Using separate increased sea surface temperature and CO2 experiments, both the rapid response cloud contribution to forcing and temperature-mediated cloud feedbacks are analyzed. Under increased temperature and CO2 changes, temperature-mediated cloud radiative feedback dominates over the rapid response in the final radiative response. Cloud feedback is positive in both long and short wave, with short wave dominating global values. Contributing most to this are low to mid-level clouds, of medium-to-high optical thickness. As a means of illustration of the methodology, a number of key parameters related to clouds, precipitation, and convection that are typically used in tuning in the model are modified. These changes result in substantial impacts on the model's current climate, but only modest changes to rapid response and feedbacks occur globally, regionally, and as a function of cloud optical thickness and height. This limited set of experiments shows that cloud adjustments and feedbacks in this model are robust under these changes, lending confidence that both model climate change projections and the conclusions of attribution studies are not overly sensitive to such parameterization tuning. Of course, a considerably larger set of experiments would be needed to demonstrate that feedbacks and rapid response are robust under the wider set of tuning adjustments commonly undertaken.

Plain Language Summary A method using a combination of sea surface temperature and CO2 forced experiments and cloud radiative kernels is used to investigate cloud feedbacks and rapid responses in the Australian Community Climate and Earth-System Simulator climate model. The methodology is demonstrated using a small set of changes to cloud parameters that are modified in a modest way as typically tuned in the model. These changes substantially alter the model climate but produce only small changes to cloud rapid response and feedbacks. Cloud feedback is positive in both long and short wave, with short wave responses dominating global values, predominantly from low level to midlevel clouds. Rapid responses are weaker in both long and short wave.