The “Index of Leading Environmental Indicators 2005” of the Pacific Research Institute says the following:
“The “hockey-stick” graph, believed to be one of the leading indicators of global warming, is now being called “rubbish.” Scientists have shown that the graph’s underlying equation would generate the same result for any series of random numbers."
In another location, a Pacific Research writer says:
Scientists agree that global temperatures have risen about 0.6 degree Celsius over the past century. But the famous "hockey stick" graph claiming to prove that the past 25 years have been the warmest in the past 1,000 years cannot be taken seriously. Experts have shown that the computer algorithm used to generate the graph would produce similar results from any series of random numbers.
The “rubbish” claim is based on a quote from von Storch, although we obviously do not disagree with von Storch on this matter. Making the reasonable assumption that the last statement of each excerpt refers to our work, neither version is exactly what we said. Ross has sent them an email me notifying them of our actual statements. In connection with an article, I re-visited MBH simulations and did some simulations, which confirm that the generation of hockey-stick shaped PC1s carried into hockey-stick shaped reconstructions. The hockey stick shape is somewhat attentuated, but the degree of attenuation in the simulations almost exactly matched the attenuation between the MBH98 NOAMER PC1 and their reconstruction. Very high spurious RE levels carry over. more
Climate Audit 
16 Comments
So this von Storch ran the PCA analysis and the hockey stick appeared – can I see it?
Also the fixation on the hockey stick being false is surely not conlusive that GW is not happening. As far as I know there are plenty of peer reviewed studies and data that show a warming trend and the hockey stick is but one of the datasets.
Ender,
You are correct on both points. However…
With regard to your first point, the hockey stick is also not proof that GW is anthropogenic, yet it is often used to promote it as such because it gives the appearance that the only demonstrative fluctuations in global temperatures are due to man’s influence.
With regard to your second point, I think that if you visit this site frequently, you will find that Mann’s hockey-stick is the graphic of choice for those promoting the idea of anthropogenic global warming. The fixation came not from skeptics thinking the hockey stick was false, but from promoters of the idea of AGW wanting to explain away the idea that recent warming was not only partly/mostly natural but that the world has been warmer before without catastrophy. Prior to the hockey stick, many graphics showed the Medival Warm Period and Little Ice Age as being extremes that could explain away much of the 20th century warming as natural variability, that the world is cooler today than in other periods in the past 1000 yrs, etc. You will find little-to-nothing in the other studies and data to explain these away. But the hockey stick erases them from existence and relegates them to regional variations.
I invite you to delve into this site. I don’t think it will take you long to realize why there is such a fixation on the hockey stick.
It certainly comes as no surprise to me that the hockey stick carries through to the temperature reconstruction. The hockey stick true fingerprint is a bandwidth-limited step function, with the step at the start of the training period (1902) and the Mann temperature curve looks more like this than it does any temperature record from the last hundred years or so. Also the cause of the promotion of the hockey stick into PC1 also results in an amplification of the variance explained, for the reasons I outlined in my weblog-breaking response to John Hunter in Lost Cedars #2.
Doesn’t look so hot for John Hunter’s hypothesis now, does it? Hopefully Steve can now get back to his broader work after this frustrating annoyance. The hockey stick is dead, QED: lets move on to other studies.
Spence_UK (posting 3): I do not have a “hypothesis” — i’m just trying to sort out what is right and what is wrong. When I see claims by the Pacific Research Institute which are clearly not supported by anything that Steve had said or done at the time (as now acknowledged by Steve), and when I read similar claims made by posters this site, I address such issues. Now that Steve has finally started to actually generate reconstructions from “random data” (after much tedious prodding from me), perhaps the technical debate can proceed based on these actual experiments.
I read the article that was in the link “What is the Hockey Stick debate all about”. In it the reconstructed data without bias shows the max temperature rise at about 1450 is 0.3 degrees (Fig 8). Now to my mind even if you accept that the data analysis is flawed the degree of heating at this time is less that half the generally accepted warming of today (0.6deg). This would tend to lend to add less weight to the idea that the Medieval Warm Period did not produce a catastrophe so global warming is not a problem. Already the modern warming is over twice this period’s warming figure and still climbing.
Also who is to say that the 1450 warming period was entirely natural. The forests of Europe were de-forested extensively at this time. In fact the rise of coal for steel making was primarily driven by lack of wood(charcoal). The effects of humans on the climate should not be minimised just because we regard them as primitive. 40 000 years ago or so the Australian Aborginies changed the face of Australia with nothing more than fire and spears.
Mr Hunter, can I suggest you try your “no hypotheses but sort out what is right or wrong” to Realclimate and try challenging Gavin & Co about the way other sites use their studies and data ? Failure to do so so will of course reveal a hypocrisy of breathtaking extent will it not ?
Ender, ofyour quoted figure of 0.6 degrees for the rise in the 20th century, at least half of that is already accepted by the IPCC as natural (that is, all the warming up to the late 1930’s), and the attribution of the rest to anthropogenic causes is purely driven by political considerations as the evidence for natural causes is as strong for the period from the 1940’s as it is for the pre 1940’s. I do like the attribution of the 15th century warming to anthropogenic causes as well, it shows an active an fertile mind, alert for every opportunity, are you sure there wasn’t a totally undetected CO2 spike caused by so far well hidden 15th century olive oil consumption peak ?
John Hunter said
You see, John, this is what scientists call “a hypothesis”. A theory, or conjecture, which is as yet unproven. I also had a hypothesis, which had rather more detailed and technical rationale behind it, outlined in Lost Cedars #2. It seems from Steve’s assessment that your hypothesis is not standing up to scrutiny, and whilst it wasn’t entirely clear before, the comments on this site that suggest the hockey stick imprint appears in the final “temperature” reconstruction can stand.
The Pacific Research website needs some tweaking but I don’t see this as strictly Steve’s responsibility, any more than I believe that Prof Mann is responsible for checking every website that uses his graph.
Nice choice of words, very apt 🙂 I would have worded it differently though; perhaps something along the lines of “wasting a lot of Steve’s time demonstrating something that the rest of us had already guessed”
As I said earlier, and I think from your comment you broadly agree, now this is settled we can move on. The hockey stick carries no weight as a temperature reconstruction. Now can we please move on to other studies!!!
Ender, if the study is flawed, the degree of heating in 1450 could have been 0.6, 0.9, or -0.3, or something else, so we cannot conclude it is less than half the “generally accepted” warming of today because we don’t know what it was.
Re#5 – Ender,
I believe the figure you refer to is Fig 8 (the number and close paraenthesis in your post shows-up here as a smiley face), which starts out at 1400 AD. You also refer to the Medieval Warm Period and the temperature differentials on the figure, as if the time periods coincide.
I don’t think you’ll find many people claim the MWP lasted beyond 1400. Many accounts say it actually ended roughly 1200-1300 AD. So if the figure you’re referring to starts at 1400 AD, it doesn’t include the MWP at all (or, at the most, the very tail end of it), and temperatures were likely higher prior to that. If you believe the figure used by the IPCC in 1995 (Figure 3), the MWP temps peaked around 1200. Figure 4 implies that temps likely peak later than shown in Figure 3 but are still headed downward before 1400. So the 0.3 degrees you speak of is small than the peak of the MWP. How much smaller, we don’t know.
Also, that 0.3 degrees you speak of is a differential with respect to a mean. The 0.6 degree change of the 20th century is relative to the beginning point of the century, not a mean. The temperature at the beginning point, 1900, is well below the mean used to center the graph. So you’re comparing apples-to-oranges with your 0.3 vs 0.6. When compared to the same mean temperature in what you called “reconstructed without bias,” the temperatures in the early 1400s were higher than those of the late 20th century. And as stated earlier, the early 1400s were likely not the warmest period of the MWP.
Keep in mind also that Figure 8 is an attempt to correct flaws in the hockey-stick methodology interpretation of proxy information. It does not attempt to account for flaws in proxy data, the possibility of which has been raised elsewhere on climateaudit.
Lastly, and what is possibly most important of all, is that the reconstructions in Figure 8 lack statistical significance. Without statistical significance, drawing concrete conclusions from the data is next to impossible. In many cases, it’s actually not even publishable.
Michael – Fair enough however then you say “Lastly, and what is possibly most important of all, is that the reconstructions in Figure 8 lack statistical significance. Without statistical significance, drawing concrete conclusions from the data is next to impossible. In many cases, it’s actually not even publishable.” Ok so we are not drawing any conclusions from the data does that mean we are not drawing the conclusion that the MWP period was warmer than today?
I don’t really get it – the data that you presented is solely so prove Mann et al wrong but you can’t draw any conclusions from it – sorry I must be really thick – I just dont see this part.
Also the 0.3 deg is with respect to a mean – what mean? – that was not clear on the graph.
Mr Snack – I am not sure what you mean. If the IPCC acknowledges 0.3 degrees as natural then it still leaves 0.3 deg as anthropogenic. You cannot discount the effects humans have on the climate. Even in the 15th century enormous tracts of land were cleared for timber and farming and to produce charcoal for smelting iron. Land clearing affects transpiration rates and can alter rainfall and so on.
Ender,
I agree, you can’t make the conclusion the MWP was warmer than today based on Fig 8. But it wasn’t meant to – it was simply done to show what Mann’s results would’ve looked like using proper computational methodology. Whether you should be able to draw any concrete conclusions from it due to its statistical nature is important to many scientific and mathematical people, but the vast majority of people just look at the hockey stick and don’t know/care about the statistics. When the hockey stick shows-up in a school book, Time magazine, an Earth Day pamphlet, a memo to a politician, etc, do you think there are any footnotes regarding statistical significance? Do you think the vast majority of readers know enough about statistics to even care?
IMHO, we’ll never come to a significant conclusion regarding the MWP. The proxy information is too scarce and likely too imprecisely correlated with temperature on a global scale. Much of what we have concerning the MWP is anecdotal information, and even this is sporadic. There can certainly be regional conclusions that the MWP was warmer than today, but on a global scale, it may be impossible to prove one way or the other.
Whether the MWP was actually warmer than today is also irrelevant, IMHO. But acknowledging it’s presence and approximating its magnitude goes a long way in determining how anamolous the 20th-century warming actually was and how much natural variability we’ve seen in the past 1000 yrs. The hockey stick implies the MWP basically didn’t exist (i.e., MWP evidence is simply regional variability), which makes the 20th century appear to be without question due solely to human GHG emissions and warming at an unprecedented and alarming rate. Most any layperson who looks at it and accepts the hockey stick to be true would easily be swayed. So I think you know why it appears in so many locations and has become the object of such fascination on both sides!
I don’t remember exactly what the data was zeroed to in Fig 8. It’s probably mentioned on this site somewhere or in Mann’s paper.
Re#11, there is no doubt humans affect weather on a local scale, and it’s possible there are climatic effects. But it’s difficult to tell how significant these climatic effects are.
Yes, the IPCC acknowledges half of the 20th century warming to be natural – the warming which occurred in the 1st part of the century. That leaves the other half, but the IPCC doesn’t say ALL of that other half is anthropogenic. I don’t remember their exact wording (which also may be different in the Working Group vs Summary for Policymakers), but I think the IPCC stated that it was either likely or certain that some of the 20th century temperature rise was due to human activity. I don’t think it was ever quanitified. It could be 0.3 deg, or it could be 0.003 deg. Nevertheless, if the warming of the 1st part of the century is entirely natural, and if it is equal in magnitude to the warming of the 2nd part of the century, then one could rationalize that a large part (or almost all) of the warming of the 2nd part of the century was also likely natural.
Michael – thank you for that considered reply. You really need to tell others that the MWP data cannot be relied on. I have seen many other sites claiming that the MWP warming is proof that the recent warming (as long as you accept it) is natural as well. Additionally there is enough doubt in the data to conclude that maybe the MWP was not as warm as we think and that maybe recent warming is significant.
Unfortunately the only real answer we will get is if the globe warms in the future. Until then, as you say, we do not really know. I really hope that you are right.
Ender,
Much of the local, historical evidence for the MWP indicates (on a regional basis) pretty consistently temperatures higher today. Of course there are always the occasional exceptions, as regional temperatures vary more greatly than global temperatures, but the balance of historical evidence suggests that the MWP was warmer than today (this evidence is easy enough to locate on the internet – check the “medieval” category for a handful of examples)
So my viewpoint would be that, on the balance of evidence available today, that the MWP was indeed a little warmer than today. The Holocene optimum was warmer still. However, this remains a “balance of evidence” rather than a clear cut proof, so will always remain a degree subjective and open to interpretation.
The multi-proxy studies were supposed to resolve this question, but in truth they have created more questions than answers.
The “pre-Mann” temperature curve produced by the IPCC back in 1995 (I think?) based on the subjective evidence shows the MWP warmer than today. The Mann curve resulted in a seed change in IPCC “opinion”.
Re#13,
I simply said I don’t think we’ll ever conclusively know whether the MWP was warmer globally or not, so the doubt goes both ways. There is plenty of evidence to suggest that the MWP existed and resulted in warmer global temps than those of today. IMHO, there is enough evidence to assume that it was warmer.
The MWP’s mere existence, regardless of whether or not it was warmer than today, can still be used to support the idea that a large portion (if not almost all) of the post 19th-century warming is natural variation.
Re ‘#15. ‘I don’t think we’ll ever conclusively know…’…’there is enough evidence to assume…’ which sounds kind of, err, conclusive ;). It’s one or the other…