Background
The authors write that "stratocumulus (Sc) cloud cover is a persistent feature of the subtropical North and South Atlantic," and that "it is well known that Sc cloud cover increases with decreasing temperatures of the underlying sea surface and that an increase in cloud cover will cool the surface temperatures via increasing the local albedo, otherwise known as the Sc feedback."

What was done
Among other things, Evan et al. state that real-world observations were used to "quantify the magnitude and spatial structure of the Sc feedback in the tropical-extratropical Atlantic Ocean," as well as to "investigate the role of the Sc feedback in shaping the evolution of coupled modes of variability there," especially when utilizing CMIP3 models.

What was learned
The four researchers report that "most models have negative biases in the mean state of Sc cloud cover and do not reproduce the observed spatial structure of Atlantic Sc clouds." In addition, they indicate that "while the majority of models exhibit some agreement with observations in the meridional structure of the Sc feedback, the vast majority of models underestimate the dependence of Sc cloud cover on the underlying sea surface temperature."

What it means
Once again, we have a situation where important aspects of both cloud type and cloud cover are simply not portrayed to an acceptable degree of real-world faithfulness in the "majority" - or better make that the "vast" majority - of CMIP3 models.