The “hockey stick” reconstruction of temperatures of the past millennium has attracted much attention – partly as it was high-lighted in the 2001 IPCC report as one of the important new results since the previous IPCC report of 1995, and partly as it has become the focus of a number of challenges. Discussion about the “hockey stick” is conducted with considerable fervor in the public media, where this curve is often presented as if it were a proof, or even the most important proof, of anthropogenic influence on climate.
As someone who has not worked on the past millennium, I do not want to discuss the merits of the often rather technical challenges (which have been dealt with elsewhere on this site). Rather, I want to discuss the “what if…” question: what if really some serious flaw was discovered in the “hockey stick” curve? What would that mean?
So let’s assume for argument’s sake that Mann, Bradley and Hughes made some terrible mistake in their statistical analysis, so we need to discard their results altogether. This wouldn’t change our picture of the last millennium (or anything else) very much: independent groups, with different analysis methods, have arrived at similar results for the last millennium. The details differ (mostly within the uncertainty bounds given by Mann et al, so the difference is not significant), but all published reconstructions share the same basic features: they show relatively warm medieval times, a cooling by a few tenths of a degree Celsius after that, and a rapid warming since the 19th Century. Even without Mann et al, we’d still be stuck with a “hockey stick” type of curve – quite boring.
So let’s try some more exciting “what ifs”. In mid-20th Century, medieval temperatures are exceeded in all the reconstructions, hence recent (last 10-15 years, say) temperatures appear to be unprecedented for at least a millennium (that even holds for the alternative histories presented by the “hockey stick” critics). Now what if that were wrong – if all proxy reconstructions as well as model simulations of the past millennium were fundamentally in error?
Let us assume that medieval temperatures after all had been warmer than the present. Even that would tell us nothing about anthropogenic climate change. The famous conclusion of the IPCC, “The balance of evidence suggests that there is a discernible human influence on global climate”, does not depend on any reconstruction for the past millennium. It depends on a detailed analysis of 20th Century data. In fact, this conclusion is from the 1995 IPCC report, and thus predates the existence of quantitative proxy reconstructions like the “hockey stick”.
Climate changes can have several different reasons, and the cause of any particular climate change needs to be investigated on a case by case basis. It cannot be found by looking at one temperature curve. Had medieval climate been warmer than the present, this would probably have been due to some natural cause – perhaps a peak in solar output. That would only tell us that in principle, natural causes can cause warming larger than what we’ve seen in the past decades. But we know that already – one need only go back far enough in time (e.g., fifty million years) to find examples of unquestionably warmer climates than today. However, it would be naive to conclude that the observed strong 20th Century warming therefore also must have a natural cause.
Investigating the cause of 20th Century warming is done in so-called detection and attribution studies, which analyze the various forcings (e.g., solar variations, greenhouse gases or volcanic activity) and the observed time and space patterns of climate change in detail. These studies, with a range of different techniques, have invariably concluded that the dominant cause of 20th Century warming is man-made greenhouse gases.
In the spirit of this article, let’s assume these studies were also wrong, in addition to all of the above. Let’s assume these studies somehow greatly underestimated natural variability in the climate system, so that the “signal” of anthropogenic climate change has not yet emerged from the “noise” of natural variations (i.e., the above-cited “discernible human influence” had not been detected after all). Surely, then we wouldn’t need to worry about global warming, and the world could hold off with the Kyoto protocol?
Unfortunately, that also doesn’t follow. The only thing that would follow in that case is that our data are not yet good enough to prove that anthropogenic climate change is already happening. That would not be so surprising – the expected amount of anthropogenic global warming to date (based on the radiative effects of the greenhouse gases and aerosols emitted by humans thus far) is only ~0.5 ºC. It is a small signal that is not easy to detect amongst the natural variability; most of the anthropogenic warming is still to come (the point of conducting science is to give an early warning, rather than just wait until the facts are obvious to everyone).
The main reason for concern about anthropogenic climate change is not that we can already see it (although we can). The main reason is twofold.
(1) Carbon dioxide and other greenhouse gases are increasing rapidly in the atmosphere due to human activity. This is a measured fact not even disputed by staunch “climate skeptics”.
(2) Any increase in carbon dioxide and other greenhouse gases will change the radiation balance of the Earth and increase surface temperatures. This is basic and undisputed physics that has been known for over a hundred years.
But how strong is this warming effect? That is the only fundamental doubt about anthropogenic climate change that can still be legitimately debated. We climatologists describe this in terms of the climate sensitivity, the warming that results in equilibrium from a doubling of CO2. The IPCC gives the uncertainty range as 1.5-4.5 ºC. Only if this is wrong, and the true value is lower, can we escape the fact that unabated emissions of greenhouse gases will lead to the warming projected by the IPCC.
Chances for that are not good. A new large uncertainty analysis that appeared this week in Nature shows that it is very difficult to get a climate sensitivity below 2 ºC in a climate model, no matter how one changes the parameters. And climate history, with its Ice Ages and other large changes, also speaks strongly against low climate sensitivity.
Discussion about the temperature evolution of the past millennium will no doubt continue in the coming years. The most fundamental problem – the sparseness of data – will not be fixed quickly, but eventually better reconstructions with smaller uncertainties will become available. However, this discussion needs to be conducted in a sober and unexcited manner; it does not help to overburden the “hockey stick” with symbolic meaning. In some media reports, the “hockey stick” has even been hyped as “a pillar of the Kyoto protocol” (which was agreed in 1997 and thus predates it) or as “proof that humans are warming the Earth”. This is a serious misunderstanding of the scientific meaning of these data.
The discussions about the past millennium are not discussions about whether humans are changing climate; neither do they affect our projections for the future. In fact, if humanity takes no action and this century will bring a temperature rise of 2 ºC, 3 ºC or even more, the current discussions over whether the 14th Century was a few tenths of a degree warmer or the 17th a few tenths cooler than previously thought will look rather academic.