Scafetta and West reply: The purpose of an opinion piece is to stimulate debate, a function quite different from that of a research paper. However, for this topic, our piece required more than the usual scientific infrastructure to avoid being dismissed out of hand.
The letter writers, whose opinions are all important, tend to fall into three categories: Peter Foukal attempts a scientific critique of our research; Roger Cohen appears to have closely followed our research and has correctly understood its scientific framework; and the other readers do not seem to know the scientific details of the debate and question our work more from a political or ethical perspective. We limit our response to cover those issues we see as crucial.
Diedrich Schmidt argues that our research is politically tainted because it was partially supported by the US Army Research Office. That charge is as ridiculous as it is insulting. We speculate, without knowing Schmidt, that he would not level the same accusation at scientists reaching conclusions with which he is sympathetic, regardless of the funding source. In America’s scientific community, there is freedom to investigate all sides of controversial issues without fear of political pressure from a funding agency.
Claiming that climate is sensitive to solar variability is not a message of despair, in part because solar activity is forecast to decrease during this century. Regardless of what the Sun may do, however, the scientific issue is what matters, and science must have priority over political correctness in this discussion.
Foukal claims that we are neglecting volcanism’s contribution to climate change and cites the important work of Gabriele Hegerl and coworkers (see Foukal’s reference 2). We did not neglect that effect. The green curve in our figure (page 51 of the Opinion piece) represents the temperature signal after the volcanic signal has been removed using a technique essentially equivalent to that of Hegerl and coworkers but applied to short time scales. The model simulations (blue and red curves), mistakenly interpreted by Wim Klaassen, have two inputs: the phenomenological signature of the 11-year solar cycle on the temperature, obtained after removing the volcanic signal, and a characteristic response of climate to external forcing of about a decadal time scale, which is essentially what many groups’ energy balance models assume.
Hegerl and colleagues attempt to interpret some paleoclimate temperature reconstructions from AD 1000 to AD 2000 by means of a simple multilinear regression analysis constructor, which is essentially a linear fit of more than one variable to the data; in their fit are four components: the Sun, volcanoes, a greenhouse gas plus aerosol component labeled as anthropogenic, and noise (see their equation 2 and table 2). The multilinear fitting coefficients, as expected, strongly depend on the particular paleoclimate temperature reconstruction that is adopted.
A careful reader of that table would notice that for some paleoclimate temperature reconstruction, the fitting parameter referring to the solar forcing is negative. That means that if the paleoclimate temperature data and the model of Hegerl and coauthors are correct, every time the solar irradiance increases, the climate cools, and every time the solar irradiance decreases, the climate warms. That looks quite unphysical. At the other extreme, if another paleoclimate temperature reconstruction is adopted, 1 the multilinear fits give very different results according to the time period fitted, and the anthropogenic component vanishes when the temperature record is fitted from the years 1001 to 1925.
One crucial difficulty surrounding the climate change issue is the huge uncertainty in the paleoclimate temperature and solar reconstruction data and in the climatic effect of the forcing. For example, according to the 2007 Intergovernmental Panel on Climate Change (IPCC) report, a doubling of carbon dioxide would increase the global temperature between 1.5 K and 4.5 K—not a small range. The uncertainty makes quite problematic the application of the traditional climate model methodology, since it relies on having correct data and sensitivities and on knowing the correct physical mechanisms. Unfortunately, those data, sensitivities, and knowledge are not currently available.
For example, when the traditional climate model simulations are compared with the patterns observed in the temperature data, models significantly underestimate the solar signature: The 11-year solar cycle signature on the surface temperature has a peak-to-trough amplitude of 0.1 K, while the climate models predict an amplitude of 0.035 K (see the Goddard Institute for Space Studies ModelE simulations at http://www.giss.nasa.gov/tools/modelE). Thus the models appear to be missing important solar–climate linking mechanisms.
Arguing, as Foukal does, that the total solar irradiance and UV radiation do not have enough power to explain the strength of the link suggests that the physics of the solar–climate linking mechanisms should be further investigated, not that the mechanisms do not exist. To think otherwise would be a logical fallacy of scientific reductionism.
As Cohen correctly observes, one major reason the scientific community has believed that most global warming was anthropogenic is the so-called hockey-stick global temperature reconstruction, 2 which is based mostly on tree-ring data. It shows little preindustrial climate variability and significant warming since 1900. That pattern supports the theory of manmade global warming. 2,3 However, the latest studies have shown the limitation of the tree-ring temperature reconstructions and that, on the contrary, climate varied substantially in preindustrial times. 2,4 The variability suggests that climate is strongly sensitive to solar change and more weakly sensitive to anthropogenic emissions than presently estimated. Thus our research and several new findings appear to indicate that the IPCC’s conclusions need significant revision.