Some remarks on a recent paper in Nature and clouds

A new paper has been released in Nature, entitled Evidence for climate change in the satellite cloud record (press release here), which talks about cloud property variations induced by human-caused global warming. It argues that primarily because of humans and recovery from the volcanic eruptions of El Chichon and Pinatubo in 1982 and 1991 (which cooled the earth), clouds are moving higher up in the atmosphere and toward the poles. This, according to the press release, exacerbates warming further, serving as a positive feedback to the human warming. These results would also confirm the predictions of cloud behaviors in climate models, which would make those models more reliable.


Global temperature responses to a CO2 doubling if the iris hypothesis (a negative cloud feedback) isn’t and is considered in climate model simulations. From Mauritsen and Stevens 2015 supplementary information.

While this paper’s arguments may all be sound (and it’s likely that there are certain positive feedbacks from clouds), evidence also exists for negative feedbacks from clouds. In my opinion, the feedback from clouds is probably not very strongly negative or positive, because of the various possible responses to global warming. I plan to do some lists on these topics in the future but in the meantime, here are some papers on negative feedbacks and clouds:

Mauritsen and Stevens 2015 argues that if the iris hypothesis is included in climate models they can be closer reconciled with reality.

Cho et al. 2012 argues that warming produces a reduction in  clouds in the Pacific warm pool which allows more OLR (outgoing longwave radiation) to escape to space, cooling the region.

Laken and Palle 2012 argues that certain SAM (Southern Annular Mode) events which are expected to increase with further human-caused warming can produce increases in cloudiness that cool South America.

Spencer et al. 2007 argues that the transition to the warm and rainy season in the tropics brings an increase in OLR due to decreases in ice clouds, supporting the iris hypothesis.

In addition, if some of the trends in the paper result from recovery from volcanoes, then the human effect may be lower than thought.