Pierrehumbert replies: Because the stratosphere is warmer than the tropopause, emission near the center of the CO2 band increases as CO2 concentration is increased with temperature held fixed. However, the increased emission in that spike is far outweighed by the decreased emission in the wings, as is verified by all detailed calculations of radiative forcing.1 The increased emission near the center of the CO2 band primarily acts to cool the stratosphere, not the surface, and once the stratosphere comes into equilibrium, the effect of ditch-shallowing becomes even less pronounced.

The papers Hardy Granberg refers to are attempts to detect the signature of recent CO2 increase in the observed trend in spectra between 1970 and 1997. That is a formidable task, given inter-annual variability, observational errors, the small signal over such a short time period, and the problem of intercomparison between measurements taken with different satellites. The spectra discussed in those papers incorporate the influence of temperature changes over the time period and do not in any way imply that the CO2 increase has made cooling to space more efficient. One doesn’t need to be able to accurately observe the short-term trend in order to confirm that radiative transfer is being done correctly. The comparisons of present-day spectra shown in my article already amply demonstrate that.

Insofar as a thick cloud acts nearly like a blackbody in the IR, the radiation to space from a region containing such a cloud indeed becomes essentially insensitive to the CO2 concentration beneath the cloud. The reason that cloud effects do not negate the basic picture I gave in my article is that high cloud coverage is sparse in Earth’s present climate, so the CO2 radiative forcing in regions clear of high clouds still exerts a dominant warming influence. What’s more, the presence of high clouds reduces the local radiating temperature, thus making the emission to space less sensitive to tropospheric temperature. This effect increases climate sensitivity. Because cloudy regions are coupled to clear sky regions by dynamical heat transports, the presence of high clouds actually means the atmosphere has to warm more in order to accommodate a given radiative forcing than it would in the absence of high clouds. These effects are fully taken into account in climate models and in comprehensive diagnostics of radiative forcing and climate sensitivity.2 

1.
W. D.
Collins
 et al.,
Journal of Geophysical Research
111
D14
D14317
(
2006
). ;
the reader can experiment with similar calculations using the model at http://geoflop.uchicago.edu/forecast/docs/Projects/modtran.doc.html.
2.
B. J.
Soden
,
I. M.
Held
,
Journal of Climate
19
14
3354
(
2006
).