Hansen Simplified

I think that I can give a very simple explanation of just how bizarre the climate reconstruction by Jim Hansen is. The graph below shows the actual Mg/Ca values for ODP806B, which is used in his reconstruction. The orange point is a modern sample in 27.2 deg C water (a value of 4.5, higher than any values in the core.) Climatological temperatures in the Pacific Warm Pool from Levitus 1994 were 29.2 deg C, which, using typical transfer functions, would yield Mg/Ca of 5.3, much higher than any values in the core. The reason why the Mg/Ca values in the core are so low is because Mg dissolves relative to Ca between the surface and the depth of the core (2500 m).

Figure 1. Actual Mg/Ca values for ODP806B (G. ruber) together with modern observation (orange) cited in Lea et al 2000 and estimate for modern Warm Pool temperatures.

So how do you splice this series of Mg/Ca values to modern instrumental temperatures. You have to estimate how much they should be moved up and how much the series should be dilated by estimating how much Mg/Ca has dissolved. How did Hansen deal with this? In effect, he simply slid the the graph so that the last point on the graph (dated to 4320 BP) lined up the red dot.

How do we know that the last dot (from the Holocene Optimum) wasn’t warmer than the red dot? Hansen doesn’t say, but it is my understanding that John Hodgman helped with the calculations.


  1. IL
    Posted Oct 3, 2006 at 12:29 AM | Permalink

    What you are showing is that despite the graph showing what looks like (*further comment in a moment) impressive precision on relative temperatures, the confidence limit on tying the method to an absolute temperature is huge, it could easily cover 10C or more and so useless for palaeotemperature estimates.
    The further comment on the relative changes in the graph – these look detailed, how is it known that these changes are not caused by diagenetic changes in the core?

    I was also surprised to see only one referee acknowledged in the paper. Even if there are anonymous referees, they are usually acknowledged which implies that there was only one reviewer. If true, that in itself is very sloppy and as for the methodology given here, this atrocious level of rigour would not get past review in my own field, nor other scientific areas of which I am familiar. Does PNAS allow critical comments? I couldn’t see in their instructions a facility for comments on papers but this paper needs it.

  2. Pat Frank
    Posted Oct 3, 2006 at 1:43 AM | Permalink

    The Acknowledgments of Hansen’s PNAS article thanks Ralph Cicerone for reviewing the submitted manuscript.

    According to this interview, Ralph Cicerone, apart from being the president of the NAS, is an atmospheric chemist. That doesn’t seem to make him particularly qualified to review work on drill core inorganic chemistry or foraminiferal temperature proxies.

    To check this, I did a search on “Cicerone RJ” between 1990-2006 on ISI SciSearch at LANL. Forty-one articles turned up, all on atmospheric gasses; primarily organics and (pseudo)halogens. These are Cicerone’s professional specialization. Nothing at all turned up concerning drill cores, temperature proxies, or the inorganic chemistry of the alkaline earth carbonates (or any other minerals) looking as far back as 1970 (134 citations).

    Speaking as a chemist myself, it appears pretty conclusive that Ralph Cicerone is not professionally qualified to have reviewed Jim Hansen’s manuscript.

    Perhaps Dr. Cicerone relied for properly critical reviews on those who were thanked in the Acknowledgments for providing “comments” on a draft of the manuscript. “Comments” don’t sound the same as ‘critical review’ to me, though.

    Or, maybe Dr. Cicerone trusted authors, “J.H., M.S., R.R., K.L., D.W.L., and M.M.-E. [who] analyzed data,” to have gotten it right.

    But those sorts of reliance or trust are failures of duty by any reviewer.

    An irony is that the same interview linked above has him as saying that, ‘Skepticism plays an important role in the scientific method,’ and that, ‘it is important to do physical calculations and understand environmental science quantitatively.

    It appears, however, that Dr. Cicerone was neither skeptical nor quantitative in his review. Nor, crediting Steve M.’s own review, thorough.

    In that regard, it was interesting to discover that Dr. Cicerone was warning about greenhouse warming at least as far back as 1986: DICKINSON, RE ; CICERONE, RJ (1986) “FUTURE GLOBAL WARMING FROM ATMOSPHERIC TRACE GASESNature 319, 109-115.

    Here’s the abstract: Human activity this century has increased the concentrations of atmospheric trace gases, which in turn has elevated global surface temperatures by blocking the escape of thermal infrared radiation. Natural climate variations are masking this temperature increase, but further additions of trace gases during the next 65 years could double or even quadruple the present effects, causing the global average temperature to rise by at least 1C degrees and possibly by more than 5C degrees . If the rise continues into the twenty-second century, the global average temperature may reach higher values than have occurred in the past 10 million years.

    He also published, (1985) “TRACE GAS TRENDS AND THEIR POTENTIAL ROLE IN CLIMATE CHANGE” J. Geophys. Res.-Atmos. 90(ND3), 5547-5566.

    He seems to have been as positive about impending climate doom in 1988 as Jim Hansen. There may, then, have been a bit of prefatory conviction directing his outlook as a reviewer.

    There’s an interesting trend apparent here. Jim Hansen was 99% sure of AGW in 1988. By 2000 and after enormous and costly modeling efforts, the IPCC was 66-90% sure of AGW. I seem to recall someone saying that something like US$40 billion had been spent on climate research between 1988 and 2000. Now, “I’m not not a statistician“, and so Steve M. and Bender can correct my model here, but setting the amount spent on climate research to approximately US$ 0.00 in 1988, it appears likely, using the climate-science-approved linear proxy model, that we’ll reach zero climate certainty after spending about US$253 billion and that we should reach that state of perfect knowledge around the year 2045.

    Isn’t science wonderful? What was it, again, that Mark Twain said about linear models? Oh, yeah: “In the space of one hundred and seventy-six years the Lower Mississippi has shortened itself two hundred and forty-two miles. That is an average of a trifle over one mile and a third per year. Therefore, any calm person, who is not blind or idiotic, can see that in the Old Oolitic Silurian Period, just a million years ago next November, the Lower Mississippi River was upwards of one million three hundred thousand miles long, and stuck out over the Gulf of Mexico like a fishing-rod. And by the same token any person can see that seven hundred and forty-two years from now the Lower Mississippi will be only a mile and three-quarters long, and Cairo and New Orleans will have joined their streets together, and be plodding comfortably along under a single mayor and a mutual board of aldermen. There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact.

    And this precocious observation regarding current practice in climate science.

  3. James Erlandson
    Posted Oct 3, 2006 at 6:12 AM | Permalink

    You agitatin’ my dots?

  4. Posted Oct 3, 2006 at 8:16 AM | Permalink

    Steve, would that mean that after application of the Lea transfer function, which is supposed to accommodate dissolution effects, that the result was not a ‘good enough’ version of absolute temperatures, that there was a post-hoc manipulation? A slide like this is effectively a manipulation of every point in the data set when the result of the calibration is supposed to recover the actual SSTs.

  5. Steve McIntyre
    Posted Oct 3, 2006 at 8:40 AM | Permalink

    I wouldn’t say that there was further post-hoc data handling. It’s just that the Lea 2000 is an empirical transformation that has no basis in biology or chemistry. Later they tried to make it work using transfer functions which attempted to account for biology and chemistry, but they didn’t use these functions (and it appears that at least one of these attempts was erroneous.)

    So Hansen used a transformation that amounts to nothing more than what I described. They moved the end-point to match and then dilated.

  6. Steve McIntyre
    Posted Oct 3, 2006 at 8:57 AM | Permalink

    Here’s Dekens et al 2002 Table 3 which shows the estimated age of the core tops for the various cores in question. It seems odd that the core tops should be so "old" and I haven’t tried to disentangle the dating so far. For ODP806B said to be “recent dO!8”, there is data at WDCP here in which the topmost value is said to be 3 cm (presumably the same as the 2-4 cm of Dekens Table 3) and is dated to 4300 BP with a dO18 value of 28.6 which I take to be “recent” only in the sense that it is not glacial.

  7. Posted Oct 3, 2006 at 8:59 AM | Permalink

    #5 Option 1. They apply Lea’s empirical transformation then splice. If so the end points of the different series are unlikely to match precisely even if right, as the data are averaged over different time intervals.
    Option 2. They apply Lea’s empirical transformation then shift the end points. Here the end points will match precisely.
    I would think 1 is OK, not great, but defensible. Option 2 is not defensible, but its seems from examination of Hansens Fig 5 that there is a slight mis-match, so he may have done Option 1. But I am only guessing, you would have to replicate it. The errors must be much larger than indicated anyway.

  8. Steve McIntyre
    Posted Oct 3, 2006 at 9:07 AM | Permalink

    The true errors are off the scale. To make an estimate, they have to back-project the original Mg/Ca ratios without knowing anything about the dissolution history. Maybe some future scientist can examine the foraminiferal shells in microscopic detail and estimate the amount of dissolution but these guys didn’t. So one is backprojecting a measurement of 4.3 or so back to the surface – did it start at 4.9 or 5.2 or 5.5? Inquiring minds want to know. Can one say, as Hansen did, that this estimate is accurate to within 1 deg C? It is such crap.

  9. Posted Oct 3, 2006 at 9:18 AM | Permalink

    #8 Estimate true precision using the cores that were not used in the calibration (out-of-sample). Possible?

  10. Steve Sadlov
    Posted Oct 3, 2006 at 9:37 AM | Permalink

    RE: #1 – First impressions and all that …. my first impression of Hansen’s study / paper was that it had been hastily slapped together. My impression is being increasingly confirmed.

  11. Dave Dardinger
    Posted Oct 3, 2006 at 9:38 AM | Permalink

    re: #8

    It is such crap.

    So where are we in the proxy survey?

    Tree rings are doubtful in even being useful temperature proxies, and even if they do turn out ok it’s certain the BCPs make the present reconstructions useless.

    Hurricanes are interesting, but the theory is still rather primative and in any case the findings are at the limits of detectability.

    Ice cores, a least in many places have problems of attribution since it’s not at all clear that the researchers have studied changing wind patterns, for instance, to see how that would affect the sources of precipitation.

    Now it looks like the “killer” forams have their own problems (though I suppose if the AGW proponents didn’t have this tendency to overreach, and stuck to the holocene they might be ok.)

    And of course we still have the problem of Jones and UHI which may or may not ever result in a real audit of his material.

    Are there proxies I’ve missed that have been gone over here?

  12. jae
    Posted Oct 3, 2006 at 9:57 AM | Permalink

    11. There are some others: stalagtites, Sargasso sea cores, other sediment cores, treelines, …

  13. bender
    Posted Oct 3, 2006 at 10:03 AM | Permalink

    Re #12
    Except that the treeline forest assemblage work of Millar et al. (2006) puts the MWP (AD815-1350) in California at +3.2°C warmer than the present.

  14. TCO
    Posted Oct 3, 2006 at 10:22 AM | Permalink

    corals have very nice year to year correlation (not just long trends) with temp. Ice cores at the poles seem to be much better then those at the tropics. Dave, may be used the critiques of tropical ice cores that we’ve had here. I also really liked the one researcher here who had some sort of proposal for an analysis that had something to do with chemical binding of different isotopes at different temps. That sounded like a killer measurement.

  15. jae
    Posted Oct 3, 2006 at 10:30 AM | Permalink

    Isotopes in peat cellulose has also been used as a proxy.

  16. Posted Oct 3, 2006 at 10:44 AM | Permalink

    #11. Sea level as a temperature proxy. Discussed that.

  17. James Erlandson
    Posted Oct 3, 2006 at 11:55 AM | Permalink

    The relationship between Mg/Ca and water temp where the plankton live is still under study.
    Further calibration of the Planktonic Foraminiferal Mg/Ca Paleothermometer: Sediment Trap Results from the Cariaco and Guaymas Basins

    Our preliminary results indicate the Mg/Ca ratio in G. ruber accurately records the observed seasonal temperature cycle … with the estimated temperatures typically being within 1.25 degrees C of the observed values.


    From the Ocean and Climate Change Institute

    We are setting up a laboratory culturing system that will let us grow deep-sea benthic foraminifera under controlled physical and chemical conditions. We will then use this system to carry out the first culture-based calibration of the benthic foraminiferal magnesium/calcium (Mg/Ca) thermometer.

  18. Steve McIntyre
    Posted Oct 3, 2006 at 12:21 PM | Permalink

    David Lea writes:

    Hi Steve: the adjustment based on the Dekens 2002 equation is quite small at
    the location and depth of Hole 806b (~0.3 deg in SST), and, in the absence
    of time-varying dissolution corrections, only serves to change the absolute
    SST value. For that reason we chose to maintain the earlier calibration to
    be consistent with the published data in the 2000 study.

    Sincerely, David

  19. John S
    Posted Oct 3, 2006 at 2:36 PM | Permalink

    “…only serves to change the absolute SST value.”

    Which is precisely the thing Hansen was using it for. Sheesh!

  20. Jeff Weffer
    Posted Oct 3, 2006 at 9:17 PM | Permalink

    Looks like a “hockey stick”, another hockey stick that is.

  21. Nicholas
    Posted Oct 3, 2006 at 11:09 PM | Permalink

    Plus 0.3 degrees may be a small number but that’s something like 50% of the warming we’ve experienced over the last 100 years. So it certainly does not seem insigificant to me, esp. if you’re going to compare it to said globally averaged instrumental temperature series.

    Of course they probably didn’t have that in mind when they did their study, but it just goes to show you how careful you have to be when citing other people’s studies and using them for purposes such as this comparison.

  22. Steve McIntyre
    Posted Oct 3, 2006 at 11:17 PM | Permalink

    In Warm Pool terms 0.3 deg C is the total amount of Warm Pool warming in the last century according to just-published D’Arrigo et al, Paleoceanography, 2006:

    Monthly gridded SST data were extracted from the Kaplan et al. [1998] data set for the warm pool region in the vicinity of Indonesia and averaged over the area bounded by the coordinates 15S–5N; 110–160E (i.e., an average of 40 grids, Figure 1). This area was selected because it encompasses much of the warm pool region associated with the Indonesian Low (a center of action of the ENSO system), while also optimizing the signal between the SST and proxy data. …Warming is evident over the past century in both the instrumental and reconstructed warm pool series (Figure 3a, both series increase by 0.2 deg C/100 years over the 1856–1992 period). Similarly, warming has been observed for Indonesia [Harger, 1995] and over the tropics (30S–30N) as a whole [Wilson et al., 2006] (Figure 3c), although over the same period, the rate of increase of mean tropical annual SSTs is marginally greater at 0.3 deg C/100 year

  23. DAV
    Posted Dec 28, 2006 at 3:31 PM | Permalink

    FWIW: I don’t know about anyone else but I can’t see any of the figures on this page or any of the Hansen posts. When I tried accessing them directly, I get a “Forbidden Access” message. Without the figures, it’s hard to follow some of the text.

  24. McCall
    Posted Dec 28, 2006 at 5:24 PM | Permalink

    re: 12 “… Sargasso Sea cores …”
    Dr Keigwin’s study included forams, as well.

  25. Steve McIntyre
    Posted Dec 28, 2006 at 7:10 PM | Permalink

    #23. I apologize about the figures. When we changed servers, most of the figures took a walk. I’m trying to track John A down to locate them.

  26. David Lewis
    Posted Jul 11, 2008 at 1:06 PM | Permalink

    Do any of you think it is wise to continue on changing the composition of the atmosphere as civilization is doing?

%d bloggers like this: