Some thoughts on Hockey Sticks

Since I’m waiting for the rain to stop so that I can finish cutting the grass, I thought I might post some thoughts I’ve had about hockey sticks. It’s partly motivated by a discussion I was having in the comments of this Quark Soup post, and partly by a sense that there is still some confusion about their actual scientific relevance. This is just going to be a bit of a brain dump, so apologies if it isn’t coherent.

  • The first thing I was going to say relates to the point that David Appell is making here. Our current understanding is that, on multi-decadal timescales, our climate responds mainly to changes in external forcings. Also, our current understanding suggests that – over the last thousand years – the largest change in forcing has been the change in anthropogenic forcing that has occured over the last hundred years or so. Therefore, we would expect the global surface temperatures of the last 1000 years, or so, to have a hockey-stick-like shape. This isn’t to say that they must have, simply that our current understanding suggests that this is the shape we would expect.
  • We have an instrumental temperature record that goes back 130 years or so. It’s highly unlikely that any proxy reconstruction will overturn the instrumental temperature record. If a reconstruction does not have a blade, or if the blade is not robust, it does not bring the instrumental temperature record into question. It might bring the reconstruction into question, but that’s a different issue.
  • Related to the above, the interesting thing about proxy reconstructions is not really the 20th portion, but the pre-industrial portion. What was our temperature history prior to the period for which we have instrumental temperatures? That allows us to put the 20th century into the context of the last 1000 years, or more.
  • Steve McIntyre (who was involved in the discussion on David Appell’s blog) seems to be highlighting that the recent Ocean2K reconstruction does not have a blade. Well, the data appears to end in 1900 and the paper title is Robust global ocean cooling trend for the pre-industrial Common Era, so why would we expect there to be a blade [edit : What I say here isn’t strictly correct. As Sou points out the paper uses 200-year bins, so the final bin is 1800-2000, but the data point shown – at 1900 – is the average of 1800 – 2000]. It does discuss the 20th century in the Supplementary information, but only considers 21 of the 57 proxies and says treat with caution. So, this paper seems to be suggesting that there was a robust global cooling trend in pre-industrial sea surface temperatures. The instrumental temperature record tell us that sea surface temperatures have been rising since the late-1800s. Overall, a hockey-stick-like shape.
  • There is little that proxy reconstructions can do to overturn the basics of anthropogenic global warming (AGW). They are, however, interesting for many reasons. For example, they can be used to try and constrain climate sensitivity (how sensitive are we to changes in external forcing) and they can be used to try and determine the magnitude of internal variability.
  • What if proxy reconstructions show large variability prior to the industrial period? Could this overthrow the fundamentals of AGW? Not really. It could mean that our climate is very sensitive to small changes and that climate sensitivity is high. It could mean that we should expect large variability on top of the long-term anthropogenic trend. It doesn’t provide a reason for questioning AGW.
  • Related to the above, those who like (or think) that climate sensitivity is low, often seem to use these reconstructions the wrong way around. For example, they argue that there was a Medieval Warm Period and that – hence – our current warming isn’t unusual. Well, the current position is that the Medieval Warm Period was not global in extent and probably not as warm as it is today. However, even if it was, it would be more indicative of a high sensitivity to small changes in forcing, than evidence against AGW. It’s certainly not evidence that we shouldn’t take our current warming seriously. If you want to argue for low climate sensitivity you should really be promoting reconstructions that should little variability, not those that show large variability and past warm periods.

Anyway, I think the rain has stopped, so I should go and finish cutting the grass. As I said, this was just a bit of a brain dump. I may well not have explained some things as clearly as I would have liked, and may have left some things out. Feel free to correct me, or add more, through the comments.

Advertisements
This entry was posted in Climate change, Climate sensitivity, ClimateBall, Science, Steven McIntyre and tagged , , , , , . Bookmark the permalink.

37 Responses to Some thoughts on Hockey Sticks

  1. jsam,
    Thanks, your comment has done so 🙂

    The interesting thing that came out of the the discussion on Quark Soup was the similarity between a figure apparently shown by Steve McIntrye at the Dec 2004 AGU meeting and a figure in the Wegman report. If you follow the link that say Postcript version here and compare it to Figure 4.4 in the Wegman report, they look essentially identical. I had thought the Wegman report was meant to be independent, especially as the first sentence says

    The Chairman of the Committee on Energy and Commerce as well as the Chairman of the Subcommittee on Oversight and Investigations have been interested in an independent verification of the critiques of Mann et al.

    Yet, what appears to be a fairly crucial figure appears identical to one that Steve McIntyre showed at a conference a year earlier, and appears not to be credited to Steve McIntyre.

  2. BBD says:

    ATTP

    Everything you say above seems reasonable enough to me, not least because I have advanced several of those arguments in the past myself so either we’re both a bit clueless or great minds etc. 😉

    Interesting that the tireless SM is making the same fake controversy argument about the Ocean2K blade that he tried on with Marcott et al. And God did that create some confusion and an unpleasant smell. He really does have a thing for millennial reconstructions, doesn’t he?

  3. BBD,
    The irony about Steve’s position with regards to Ocean2K and Marcott et al. is that he seems to have criticised Marcott et al. for including a blade that was not robust (as admitted in the paper) but is now criticising Ocean2K for not including a blade that would not be robust if included.

  4. BBD says:

    It’s nuts. O2K ends in 1900, as you point out above. SM’s making a big deal about it being deliberately kept out of AR5 because not bladed and so not on-message is truly silly (and a conspiracy theory to boot).

    It’s as if he expects the majority of his readership to be too daft to notice that the reconstruction stops when it does. And too daft to ignore the post-1900 sea level data and the OHC reconstructions from ~1950 through to ARGO to the present.

    It’s transparent, clumsy mischief-making with exactly no substantive basis, just like with Marcott, as you say.

  5. BBD says:

    Oh, and my father always said that you shouldn’t cut the lawn when it was wet. Presumably because it rips it up or some such. 🙂

  6. BBD,
    My particular concern is related to my lawnmower being electric 🙂

  7. BBD says:

    Oh go on, live a little… 😉

  8. The main interest in the hockey sticks is in the “shaft”; I thought everyone knew that. Only the wackos argue about the rise during the instrumental record which, as you say, is better represented by the instrumental record. The interest is in the degree of variability in the pre-instrumental period. The original HS was quite flat during that period.

  9. William,

    The main interest in the hockey sticks is in the “shaft”;

    Yes, exactly.

    The interest is in the degree of variability in the pre-instrumental period. The original HS was quite flat during that period.

    Yes, and the newer reconstructions are not as flat, which many seem to argue suggests that MBH98 was flawed, rather than simply one of the first.

  10. Sou says:

    I don’t know that the recent ocean2k paper ended in 1900. I think what it did was end in the 1801 to 2000 “bin”, which would have included the coldest years of the past 2,000 years, as well as whatever proxy records were included up to 2000. The boxes in Figure 2 showed a lot of things, including the median for each 200 year bin, the latest of which was centred on 1900 – but went from 1801 to 2000.

    It’s correct to say that it doesn’t highlight the recent warming, which we get from thermometers. No need to bin proxies. I say Steve McIntyre is a thermometer denier 😀

  11. Sou,
    I see, yes, I think you’re right. I was focusing on the Supplementary Information it talks about

    A subset (n = 21) of the Ocean2k 57 input reconstructions are dated by 210Pb dating or annual band counting, and are of sufficient resolution to potentially record 20th century SST changes

    But I see what you’re getting at. The main paper used 200 year bins, but they then included a section in the Supplementary Info that tried to use a subset at 25-year resolution to consider the 20th century SST changes.

  12. Pingback: Science's best friends - Ocasapiens - Blog - Repubblica.it

  13. Wait a minute, I thought “no one denies that the earth is warming” ;).
    Yet another example of the inconsistencies of the denial-o-sphere.

  14. guthrie says:

    I recall from my lawn mowing days that wet grass tended to lie down flat in front of the blade, whereas dry grass was comparatively rigid and cut well. Also damp grass coated the inside of the lawnmower more easily and was a pain to get rid of.

    As for the proxy reconstructions being more variable, that is surely just an artefact of their not being very accurate in the first place, and tell you nothing about climate on the short term, say, decadal scales, that we have been able to reliably record for the last century and more.

  15. Gator says:

    ATTP — the Wegman report was a sham. Read through that deepclimate link and also this: http://www.desmogblog.com/see-no-evil-speak-little-truth-break-rules-blame-others
    John Mashey has been very thorough in documenting the mess that is Wegman and Said.

  16. Put simply, the handle of the hockey stick is natural so, as those in denial like to remind everyone, has always changed. The blade is all us, but is still attached to a fluctuating base line.

  17. anoilman says:

    Around here, we use the ‘drought’ method of preventing lawn growth;
    http://globalnews.ca/news/2179582/alberta-declares-disaster-after-losses-from-drought/

    Of course, we’re getting some rain now which will really bring up our yearly average. The fact that its at the end of the agricultural cycle will be completely lost on many of those in the denial sphere.

  18. dave says:

    …and Then There’s Physics says: [September 13, 2015 at 2:44pm]
    “Yet, what appears to be a fairly crucial figure appears identical to one that Steve McIntyre showed at a conference a year earlier, and appears not to be credited to Steve McIntyre.”

    This isn’t quite right, a point I’ve also now made at David Appell’s Quark Soup. While McIntyre’s “Jan. 5, 2005 Postscript Graphic” set of 12 “hockey sticks” looks remarkably similar to the Wegman Report’s Figure 4.4 from 2006, they’re not identical as the individual “hockey sticks” differ. Also, they called Figure 4.4 “One of the most compelling illustrations that McIntyre and McKitrick have produced”, so they did give them credit, in a way.

    Deep Climate’s “Replication and due diligence, Wegman style” blog post of November 16, 2010, discussed how Wegman’s illustrations were rendered directly from McIntyre and McKitrick’s archive which selected the 1% most pronounced “hockey stick” PC1s. Further on, Deep Climate “turned to Fig 4.4, which presented 12 more simulation PC1 hockey sticks. Although this figure was not part of the original M&M article, there is a fourth figure generated in the script, featuring a 2×6 display, just like the Wegman figure. A quick perusal of the code shows that these too were read from McIntyre’s special 1% collection, although a different selection of 12 PC1s would be output each time.”

    So, my guess is that McIntyre used that script to produce the “2005 Postscript Graphic”.

    A year and a half later, the Wegman Report said “While at first the McIntyre code was specific to the file structure of his computer, with his assistance we were able to run the code on our own machines and reproduce and extend some of his results.”

    Looks like they used the same script selecting from the same special collection to produce their Figure 4.4.

  19. mt says:

    1) We did not know how variable the GMST was on century time scales prior to publication of MBH98. The low variability made for a compelling graphic, but one could argue that the (explicit) uncertainty left room for low frequency wobbles.

    2) Immediate criticism from responsible sources suggested that the methodology underestimated the wobble. It was generally agreed that while the MBH graph told a compelling story, it was also a reassuring one – it did indicate that the system had little decadal to century scale variability and responded predictably to forcing.

    3) As you point out, succeeding efforts have been more wobbly.

    4) BEST is a notable exception, though. It shows that CO2 and volcanoes as enough to explain every bump and wiggle. This is surprising, and brings us back again to the vision of the forcing -> GMST transfer function as rather simple and basically linear.

    5) But putting my quasi-skeptic hat on, I have to wonder what sort of a millenial record they might see as reassuring. Surely not every imaginable trajectory of GMST supports the alarmonists? Nowadays it is very hard to imagine. But put yourself back to 1998. What would the quasi-skeptics have WANTED to see?

    I think in their world, the wigglier the graph is, the better, because it becomes relatively easier to delay the reckoning with the oddity of the contemporary evidence. The more bumpy, the better. This explains their bizarre attachment to the MWP. They want to believe that the system wobbles for mysterious reasons, and that despite the wobbliness, it is impervious to human disruption. This becomes a matter of faith. So the hockey stick is wrong, and Clausius-Clapeyron feedback is speculative, and who knows if the CO2 is even man made, and so on to ever shallower and stupider arguments.

    6) In our world, the wiggly system is more prone to sudden transitions. Who knows what happens once we kick the system hard enough. A reasonable delayer (if any still exist, which I doubt) would be thrilled by the hockey stick. But a reasonable delayer would also be arguing that climate models are very very good, not very very bad.

    Both the bumpy trajectory and the blind theory make the risks worse in a rational argument.

    But we are not arguing with rational people. Naysayers are grandstanding to people who are paying very little attention. We should really stop trying to help them pretend that they are making any sense.

  20. dave,
    Okay, the first time I looked at it I thought it was virtually identical, but you’re right, it’s not.

  21. dave says:

    ATTP,
    Yes, they’re so similar they look identical at first glance.
    Deep Climate was able to see McIntyre’s code, but despite Wegman asserting that scientists should make code available for “audit”, the Wegman Report code has never been released. Perhaps because that would reveal they simply used McIntyre’s script without understanding it?

  22. dave,
    Yes, I think it would be extremely difficult to show that Wegman happened to write a completely independent code that also – by chance – decided that the best way to illustrate that MBH’s methodology produced hockey sticks was to select the 100 most hockey-stick-like from the sample of 10000, rather than selecting randomly from this sample. Especially as MM05 does not make clear that their sample of 100 is not randomly chosen.

  23. John Mashey says:

    1) The key post as noted above was DC’s Replication and due diligence, Wegman style.

    2) Nick Stokes usefully added Effect of selection in the Wegman Report

    3) Code: Wegman promised it to Henry Waxman, gave a false reason for delay, never delivered the code that would obviously have been ~same as McIntyre’s.

    4) I do not use the term lightly, but the Wegman Report was fraudulent in even more ways than I knew when I wrote Strange Scholarship… (2010). See FOIA Facts series.

  24. JohnMashey says:

    1) The key post as noted earlier is DC’s Replication and due diligence… (2010), but Nick Stokes’ <a href="http://moyhu.blogspot.co.uk/2011/06/effect-of-selection-in-wegman-report.htmlEffect of selection in the Wegman Report has useful further analysis and graphics.

    2) Wegman promised code and data to Henry Waxman, but used <a href="http://www.desmogblog.com/ed-wegman-promised-data-rep-henry-waxman-six-years-ago-where-it a false claim of Navy delay.
    They never delivered, which indeed woukd have made the source ibvious.

  25. There’s plenty of time to whine about “Mr. Rice,” but no time over almost a decade to audit nonsense which posterity has been told McIntyre and McKitrick produced. Fascinating priorities.

  26. DumbSci,
    Yes, I noticed Steve’s new post. Glanced at it, but didn’t read it thoroughly. ClimateballTM 🙂

  27. Well, The Auditor managed to make it through an entire blog post without calling anyone a pet rodent, or “satirizing” climate scientists as hissy fit throwing addicts of heroin and crack cocaine. Baby steps?

    Maybe after some more baby steps The Auditor will finally audit Wegman’s Fig 4.4. I can’t wait!

  28. DS,

    Maybe after some more baby steps The Auditor will finally audit Wegman’s Fig 4.4. I can’t wait!

    Yes, especially as he didn’t seem to get any kind of credit, despite it seeming quite clear that he had some kind of hand in it’s generation.

  29. As Dave noted, Wegman did give credit by saying McIntyre and McKitrick produced the figure. That actually might be why Wegman and The Auditor refuse to answer any questions about the figure which will define their legacies and (to a lesser extent) the future of our civilization.

    Wegman claimed to be writing an independent report, so he can’t simply admit that he copied McIntyre’s code without even noticing a glaring 100:1 cherry-pick. And apparently he’s the kind of professor who can’t admit to mistaking AR(1) noise with parameter 0.2 (decorrelation time of ~1.5 years) with the ridiculous ARFIMA noise McIntyre actually used (decorrelation time of ~350 years).

    The Auditor refuses to answer questions about those flaws, perhaps because he was given credit for these misleading claims almost a decade ago and never (to my knowledge) objected. The Auditor seems to be fully aware that his legacy will be defined by this shameful figure, but he simply doesn’t care.

  30. On the 2015-09-19, at 11:48 (ADT), the Auditor writes:

    I’ll also discuss untrue claims by [AT and others] that the Ocean2K data “finishes in 1900” or is otherwise too low resolution to permit identification of a concealed blade.

    On the 2015-09-13, at 16:27 (ATDT), Sou writes:

    I don’t know that the recent ocean2k paper ended in 1900. I think what it did was end in the 1801 to 2000 “bin”, which would have included the coldest years of the past 2,000 years, as well as whatever proxy records were included up to 2000. The boxes in Figure 2 showed a lot of things, including the median for each 200 year bin, the latest of which was centred on 1900 – but went from 1801 to 2000.

    On the 2015-09-13, at 16:39 (ATDT) writes:

    I see, yes, I think you’re right. […] But I see what you’re getting at. The main paper used 200 year bins, but they then included a section in the Supplementary Info that tried to use a subset at 25-year resolution to consider the 20th century SST changes.

    AT’s post has been edited the very same day it has been published.

    The Auditor’s post has been posted 6 days after that.

    ***

    Oh, and the Auditor’s reaction to Ritson’s request was:

    In the comment thread, Ross:

    The GRL peer review process twice ensured that Ritson’s gibberish never saw the light of day, despite the Editor apparently wanting to print it. Five prime examples of the usefulness of peer review, all of which Mann rejects in favour of unpublished email diatribes from Ritson, even while claiming the mantle of peer review for himself. And to top it all off I’ll bet even Mann hasn’t been able to figure out what Ritson’s on about. The whole situation is lunacy.

    http://climateaudit.org/2006/08/31/do-climate-scientists-need-ritalin/#comment-62269

    Try to make this up. You can’t.

  31. DY,
    The caption of Figure 4.4 seems somewhat ambiguous though. I read it as crediting their analysis, rather than them for producing the figure. I guess one could argue that it does give credit, in a sense.

  32. Also note, from the archived link courtesy of Dumb Scientist, here’s a glimpse of what the Auditor learned from his AGU presentation:

    People who were not mathematically inclined were intrigued by a graphic showing 8 hockeysticks – 7 simulated and 1 MBH (the same sort of graphic as the one put up here a while ago, but just showing 1 simulation.) Quantity seems to matter in the demonstration. No one could tell the difference without being told. I’ll insert this graphic here in a day or two.

    http://web.archive.org/web/20050124025310/http://climate2003.com/

    Yup.

  33. I hadn’t realised that one of the panels in the figure that Steve posted in 2005 (that is linked to from the second comment on this post) is the MBH98 NH reconstruction. I think it’s the second one from the top, on the right. This seems to create an interesting issue. If you consider Figure 1 in McIntrye & McKitrick (2005) it shows a simulated HS in the top panel and the MBH98 reconstruction in the bottom. The y-axis range for the two panels are, however, quite different. The top panel goes from -0.08 to 0.02, the bottom from -0.4 to 0.2. I think the top is intended to be some kind of scaled units, and the bottom is deg C. The y-axis ranges on the panel of 12 (one of which supposedly contains MBH98’s reconstruction) all seem to go from -0.08 to 0.02. Would be interesting to know what scaling factor was used for the MBH98 reconstruction and why it was chosen.

  34. Willard’s 6 day delay is 4 days more amusing.

    ATTP, agreed on the ambiguity. Curious to see if they’ll try to give each other the “credit” or keep their apparent vow of lifetime silence. (Aside from telling profs to “take a Ritalin” for their “gibberish”- LOL!)

  35. I’m no dendrochronologist, but Wikipedia’s overview seems helpful. Mann, Park & Bradley 1995 didn’t scale proxy records to obtain temperature, but MBH98 was the first to use a scale factor determined by principal component analysis of the proxy records vs. instrumental record PC’s during the calibration period from 1902 to 1980.

    It would be even more interesting to know if it’s ethical to quietly change the scale (and units) of the MBH98 reconstruction graph to hide it among a cherry-picked 1% of simulations based on input noise with much longer decorrelation time than the US Congress was told.

  36. It would be even more interesting to know if it’s ethical to quietly change the scale (and units) of the MBH98 reconstruction graph to hide it among a cherry-picked 1% of simulations based on input noise with much longer decorrelation time than the US Congress was told.

    Yes, what I was trying to say, but not nearly as clearly as you’ve managed.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s