Duffy, Santer, and Wigley reply: Solar irradiance measurements have been made by a number of satellites covering different time periods. Several investigators have stitched together the multiple records into composites, correcting for small instrumental differences (for a comparison, see the online version of this letter). Nicola Scafetta and Bruce West make much of the fact that our figure showed the PMOD composite rather than their favorite, ACRIM. The differences between the two, however, are insignificant in terms of implications for climate; neither produces anything close to the observed late-20th-century warming, even if one assumes a climate sensitivity much greater than the most commonly accepted value. Furthermore, the superiority of the ACRIM composite is not established. 1  

Scafetta and West’s characterization of the 2004 paper by Gerald North and coworkers (reference 4 in Scafetta and West’s letter) contradicts that paper’s abstract. Far from finding that “the climate is much more sensitive to solar changes than what climate models predict,” North and coworkers find “a faint response to the solar cycle” with amplitude “roughly what we would expect (a few hundredths of a degree) based on simple energy-balance model estimates.” That finding contradicts Scafetta and West’s argument that the climate is mysteriously hypersensitive to solar variations.

We used a 10-year running mean in our figure 2 precisely because it masks the 11-year solar cycle; our point was that there is no significant multidecadal trend due to solar variability.

Scafetta and West’s discussion of glacial and interglacial cycles does not support their assertion that climate is exceptionally sensitive to solar variations. As is well established, glacial and interglacial temperature differences result from extremely large changes—not “small” ones as Scafetta and West claim—in the spatial and seasonal patterns of incoming solar radiation, which trigger two powerful but slow feedbacks: changes in atmospheric carbon dioxide and changes in surface reflectivity resulting from the advance and retreat of land ice sheets. Certainly, neither feedback can be responsible for late-20th-century warming.

Although this is irrelevant to the main point of contention, climate models do not assume that “only humans can modify greenhouse gas concentrations.” Naturally occurring CO2 variations are included either by prescription or through modeling of climate and carbon-cycle feedbacks.

Finally, a recent paper 2 explains in detail the serious flaws in the work of Scafetta and West. Primarily, multicollinearity between different climate forcing agents makes it impossible to unravel their relative effects by considering only a single forcing, as Scafetta and West attempt. Reference 2 further shows that the statistical method they used leads to grossly incorrect results; when applied to a situation with a known solar contribution, it gives a greatly and unrealistically enhanced solar effect.

In response to Benjamin Jordan, we note that observed temperatures reflect both natural variability and the effects of forcings such as greenhouse gases and solar variability. So in an era of increasing greenhouse gases, each year need not be warmer than the previous, even as temperatures trend generally upward. Climate models correctly predict that phenomenon. 3 However, because climate simulations are not initialized from observations in the same way that weather forecasts are, they are not expected to predict the timing of natural variations, including cooling episodes. Hence, the lack of any warming trend since 1998 is not cause for concern about climate models.

In summary, we do not claim that the climate is insensitive to solar forcing, only that the sensitivities to different types of forcing appear to be very similar. We are open to the possibility that unknown feedbacks might amplify solar forcing; however, Scafetta and West have provided no evidence of such and no reason to discard an explanation of late-20th-century warming that is consistent with theory, models, and observations—namely, increased greenhouse gases.

1.
See, for example,
M.
Lockwood
,
C.
Fröhlich
,
Proc. R. Soc. A
464
,
1367
(
2008
).
2.
R. E.
Benestad
,
G. A.
Schmidt
,
J. Geophys. Res. D
114
,
14101
(
2009
),
.
3.
See, for example,
D. R.
Easterling
,
M. F.
Wehner
,
Geophys. Res. Lett.
36
,
L08706
(
2009
),
.