Comment op all wrong (Score 3, Informative) 249
the abstract doesn't say they used data, it says they identified a math procedure that caused variation between the models
so, what you have are a lot of complex computer models that vary in output; the authors show that about half the variation is due to cloud mixing
however, we have no idea if the models are in fact accurate, other then Fig 1b of Fyfe etal, which suggests that the models are in fact NOT accurate, so it doesn't matter if you lower the variation between them.
http://www.see.ed.ac.uk/~shs/Climate%20change/Climate%20model%20results/over%20estimate.pdf
I would remind people of history: in the early 1800s, people realized that CO2 absorbs IR, and the late 1800s, they realized that humans were actually putting out enough CO2 to make a diff
Then, around 1900, someone pointed out that the atmosphere is optically thick in the IR (if you could see the color "IR" it would be pitch black all the time), so an increase in CO2 shouldn't matter
This *scientific consensus* lasted untill the 1950s, when people realized that it is emission from the outer atmosphere that matttrs....
so, for 50 years, there was a consensus that CO2 human warming was hooey