...Mark Zelinka, the first author of the study, said in an email that the only dramatic change between the two generations of models was how climate warming due to cloud cover had been incorporated. He explained that clouds reflect some sunlight that hits them, and as the planet warms from CO2 emissions, cloud cover will decrease. This in turn causes a positive feedback loop and further warms the climate.
Ceppi added that clouds also have their own greenhouse effects and can act as a blanket in the same way as greenhouse gases. The strength of this property depends on how high a cloud is, adding another layer of complexity that has only recently been incorporated into models.
Just because clouds are being modeled more realistically does not make the model as a whole better, though. In a preprint under review for the journal Earth System Dynamics, researchers at the University of Exeter looked at whether high estimates of climate sensitivity made sense in the context of historical observations.
First author Femke Nijsse said she was spurred to look into sensitivity because the last IPCC report considered a likely range to be lower than what several new models have predicted.
“We were quite surprised that this new generation of models showed quite a few models with a very high sensitivity,” she said.
Nijsse and her team found that at least six current models were inconsistent with historical climate data, likely because of the new cloud modelling.
Ceppi agreed that some of the models predicting a higher sensitivity than before may be unrealistic; this presents a “bit of tension,” since these new models were designed to better represent the climate processes that occur in clouds.
“On the one hand, they should be better, but
on the other hand, at least some of these models seem to produce sensitivities that are too high compared with observed temperature changes,” he said. “That's a bit of a puzzle there, that's something we need to resolve.”