Talk to the Hand

By email, I received anonymous notice of a complaint against GRL which seems to indicate Climategate style business as usual at their office.  This is with respect to a paper submitted in July09, well prior to climategate, a rebuttal by Foster (Tamino) and crew, and a rejected reply by the original authors. I would love to slam the team one more time, because they are what they are, however in this case they don’t appear to be wrong in their critique of the paper.  Of course correctness doesn’t excuse the incestuous behavior which seems to continue at GRL in the post Climategate world.  The boys really, aren’t listening to us in any way at all.

An article at Icecap covers most of the points.  Authors denied the right of reply.

There are several links in the discussion but the crux is, publish a paper which shows much of global warming can be explained by ocean currents, angry taminoesque critique published as a reply and authors attempted defense rejected.   What makes it interesting is that some of the behind the scenes discussion of this paper was shown in climategate emails.

First the authors have a summary of events in this pdf. McLeanetalSPPIpaper2Z-March24[1]

Here are some emails and comments on the first page of the above pdf link.

“Having now read the paper [McLean et al., 2009] in a moment of peace and quiet, there
are a few things to bear in mind. The authors of the original will have a right of reply, so
need to ensure that they don’t have anything to come back on.”
Phil Jones to Jim Salinger, July 28, 2009

Below is a critique of the tone, most of us know Tamino is a ja…… already so this isn’t surprising.

“But as it is written, the current paper [Foster et al. draft critique] almost stoops to the
level of “blog diatribe”. The current paper does not read like a peer-reviewed journal
article. The tone is sometimes dramatic and sometimes accusatory. It is inconsistent with
the language one normally encounters in the objectively-based, peer-reviewed literature.”
Anonymous referee of the Foster et al. critique, September 28, 2009

And some hint of the behind the scenes discussions which led to the critique being accepted.

“Incidentally I gave a copy [of the Foster et al. critique] to Mike McPhaden and discussed
it with him last week when we were together at the OceanObs’09 conference. Mike is
President of AGU. Basically this is an acceptance with a couple of suggestions for extras,
and some suggestions for toning down the rhetoric. I had already tried that a bit. My
reaction is that the main thing is to expedite this.”
Kevin Trenberth to Grant Foster, September 28, 2009

In order to understand the whole situation, you actually need to read the papers and critiques.  Since the critique is above, I looked for and was lucky to be able to find the paper on line here. InfluenceSoOscillation[1] as copied from this link http://icecap.us/images/uploads/McLeanetalSPPIpaper2Z-March24.pdf.

The Foster critique can be shown in the following graphs from the original paper.

The authors compare the derivative of the values from the top pane to determine the influence of the ENSO on global temps, the problem is that they only explained the short term variance and not the long term trend.  By taking a derivative, you remove the slope component from the data.  This is admitted to fully in the rebuttal portion by the authors although they state that their argument about being able to describe long trends stands.

The original paper states this:

We have shown here that ENSO and the 1976 Great Pacific Climate Shift can account for a large part of the overall warming and the temperature variation in tropical regions.

From derivative (trendless) based comparisons of temperature. An excerpt from the authors response to Foster (Fea10) is below:

Fea10 state that the method of derivatives that we employed would minimize long-term trends. We
completely agree, and wish to stress that we used derivatives only to ascertain the existence of the
relatively consistent time-lag that exists between changes in the El Niño-Southern Oscillation (ENSO)
and later changes in the global average lower tropospheric temperature anomaly (GTTA). Having
demonstrated that a lagged relationship exists between changes in the Troup SOI and the lower
tropospheric temperatures (LTT) data from satellite microwave sensing units (MSU LTT), we then
investigated whether this was a chance artefact of the data by testing (successfully) for a
corresponding relationship between the Troup SOI and radiosonde data (RATPAC-A LTT), and the
Troup SOI and Tropical MSU LTT.

I just can’t figure out how their paper has anything to do with long term trends.  From my perspective, they seem to have overconcluded a bit. I think it would have been a fine paper with a lot to say about the effects of southern oscillation on global tropospheric temperature but little was demonstrated WRT long term trend.  I also don’t see where Tamino’s critique proved there isn’t a linkage and really don’t understand why GRL couldn’t publish the authors reply.

Maybe I’m missing something.

16 thoughts on “Talk to the Hand

  1. It would seem one must account for the effects of El Chrichton and Pinatubo had on global temperatures (and the long term “trend”). Considering there was no warming in TLT through 1997, but after the 97-98 El Nino an apparent step change ensued, how can anyone conclude ENSO does not have long term effects? Was it an adjustment for the interruption caused by the volcano episodes? After the ‘Super El Nino’ event it took several years for the East Indian and Western Pacific SST to drop back to pre-El Nino SST. Why?

    Now in the current El Nino which is also releasing an enormous amount of heat, will we see a new step change but at a lower threshold? Also, if one believes the XBT OHC data previous to ARGO, OHC in the upper 750m was still increasing but stopped in 2003 and now according to the latest estimate has been in a slight cooling mode.

    So where is the CO2 warming signal in the TLT tropical data which is where its supposed to be according to the CO2 AGW hypothesis? Inquiring minds want to know.

  2. I don’t know what you are missing, Jeff. There are big issues and interests at play. Maybe Craig Loehle will comment at some point. I recall now that I stumbled across the docs you have at the icecap, from where…I don’t recall.

    My interest was that de Frietas was associated and they went through a timeline. Nobody would have ever known about this ‘peer-review’ poor gamesmanship without the release of the emails. Regardless of the absolute correctness of the McLean, et al paper they should have been allowed to publish the quick response. The team’s pile on notion to anything they don’t like reminds me more of a jr. HS debate club.

    I won’t comment on the papers directly but one problem I consistently have is, including McLean, they all try to do too much. You used the phrase “over conclude”. I call ENSO, “Enso, da’ baker”.

    Really, though, if someone would ‘figure out’ the causes of ENSO, PDO, MDAO, etc. then I would like that part…As opposed to, “The physics says…” and then play correlation poker.

    Heads up on one issue I’ve been looking at and historically related to your post. Solomon, et al in a recent Science, M. Bender, et al in Science, and another author in Science (I forget off the top of my head, maybe N2O generation in oceans) are all putting as reference 1 the IPCC AR4 report. I call that coat-tailing. Science is letting them reference the policy report. Perhaps in order to get ready for the claims that AR5 is totally based on the the peer-reviewed lit. I mention this because McLean, et al, make the point that that the ‘rebuttal’ is by IPCC authors.

  3. They could have come to conclusions about the long term trend if they had calculated the magnitude of the effect of a change of SOI on the dT. In principle the would get a coefficient that could be used on the SOI to get the T.

    My bet is they would have found that the long term trend was not explained by the SOI, at least mostly.

    Another case of denial of the right to respond to criticism:

    http://sites.google.com/site/rossmckitrick/

    In 2007 I published a paper with Pat Michaels showing evidence that CRU global surface temperature data used by the IPCC are likely contaminated due to socioeconomic development and variations in data quality. In 2008 Gavin Schmidt published a paper in the International Journal of Climatology claiming our results, as well as those of de Laat and Maurellis who independently found the same things we did, were spurious. My rebuttal, coauthored with Nicolas Nierenberg, was submitted to the IJOC in April 2009.

    McKitrick, Ross R. and Nicolas Nierenberg (2009). Correlations between Surface Temperature Trends and Socioeconomic Activity: Toward a Causal Interpretation.

    We found out in February that it has been rejected. Interestingly, it turns out that the IJOC had sent Schmidt’s paper, which focuses on defending Phil Jones’ CRU data against its various critics, to be reviewed by none other than Phil Jones of the CRU. As you can imagine the review was rather enthusiastic and uncritical. The IJOC didn’t ask deLaat or me to supply a review, nor did they invite us to contribute a response. And they have rejected the response we did submit, on the basis of some loopy referee reports to which Nico and I were not given a chance to reply (though we did anyway). Nice way they run a journal over at IJOC. The paper is being upgraded and submitted elsewhere.

  4. Hey Jeff,

    I stumbled across this at Bishop Hill’s blog earlier this week and have been wondering about it since. My biggest question in all this is: How often does the original authors response to a critique get rejected by referees, and the journal decide not to publish? Does it happen often?

  5. hah, somebody named James Lewis at PJM has a link to the Solomon paper’s abstract. I might read more. “shameless-science”.

  6. Jeff Id,

    I would love to slam the team one more time, because they are what they are, however in this case they don’t appear to be wrong in their critique of the paper. …

    … authors attempted defense rejected.

    Given that you admit that Foster et al. are most likely correct, and McLean et al. are most likely wrong, what kind of “defense” were you thinking that McLean et al. should be offering?

    Perhaps something like ‘yeah, we were wrong, but they laughed at Galileo too, so even though we were wrong, we were right’?

    Or ‘yeah, we were wrong, but we just had to find a lame excuse to scribble a reply, so we’ll just say that we were right’?

    Or ‘yeah, we were wrong, but OH LOOK WE HAS THEM STOLEN EMAILS AND A MIRACLE OCCURS HERE we were right’? (which, by the way, is what McLean et al.’s ‘response’ was basically)

    Seriously?

    An article at Icecap covers most of the points. Authors denied the right of reply.

    Wrong. McLean et al. had the right of reply, and they decided to exercise this right by writing nonsense. The reviewers simply exercised their right — nay, their duty — of quality control.

    bi

  7. Jones – ‘The authors of the original will have a right of reply, so need to ensure that they don’t have anything to come back on’.

    Well one would hope that the response would be of such a standard that McLean/Carter would not ‘have anything to come back on’.
    If, however, the authors were to be denied ‘right of reply’ could the scientific standard of the response therefore be allowed to change?

    Can you imagine Jeff Id ever saying this to Roman, or Steve Mc to Lucia…… I would bet ALL my quatloos on the temperature rising rapidly beyond a tipping point.

  8. #6, If I though that the paper was wrong in every sense, I would agree. At this point it appears to be a bit of an overreach for the paragraph pointed out.

    There are reasonable “potential” relationships which may be creating trend at the same time as they create short term variance. The link is quite strong in the short term variance of the two plots. It begs for more research to me rather than strong conclusion. It’s not like the paper was wrong, it’s more that a conclusion of being a large source of warming was not proven or disproven. The variance match is reasonable evidence that it might be the result of ocean currents though.

  9. For Jeff
    NOTE:1
    “high frequency” means: “more than other occurrence”
    NOTE:2
    derivatives are needed because temps of water and land change at different rates IE water 1c and air 10c for the same amount of heat (de trend)

    Click to access McLeanetalSPPIpaper2Z-March24.pdf

    “just can’t figure out how their paper has anything to do with long term trends. a fine paper with a lot to say about the effects of southern oscillation on global tropospheric temperature but little was demonstrated WRT long term trend.”

    in appendix “B” Page 19
    “This conclusion also
    implies that a high frequency of occurrence of El Niño conditions will cause a relatively sustained
    period of elevated temperatures, such as might popularly be termed ‘global warming’, whereas a high
    frequency of occurrence of La Niña conditions might be termed ‘global cooling’.”

    Look at it this way…. IF McLean et al (2009a) did not hit the mark THEN why did Tamino get off his but and do a rebuttal paper?
    I have to say that what I could find on him does him no good in my book.
    he is an expert on global warming because he is an astrologer? or a mining engineer?
    huh?
    Jeff, I think you are a true warmer if you do not see the McLean et al (2009a) high frequency of occurrence of El Niño
    makes for a warming of the thermometers.
    You told me because you have 1000 thermometers you can get .0001 degrees accuracy, now we are telling you that because we have 1000 El Niño’s that we have a warming trend, get it? high frequency of occurrence?
    SEND me an e-mail please so I know you at least you read this.
    I put a copy on my desktop.

    entertainment of google for info on tammy

    5-sigma confidence
    http://motls.blogspot.com/2010/03/tamino-5-sigma-and-frame-dragging.html

    http://www.citizensky.org/users/tamino

    gfoster at aavso.org

    OR
    University of Utah B.S. 1986
    – Mining Engineering???
    http://www.hollandhart.com/peopleprofile.cfm?IDName=PersonID&ID=5144

  10. I think this episode shows a major problem with climate science and peer review. The adversarial side appears to work well for those scientists in the advocacy camp for immediate AGW mitigation while the other side is apparently shut out of the process. I can see valid criticism of the McClean paper, but where was a similar response published for much of what passed peer review for Mann et al. in their proxy reconstructions and otherquestionable papers that have been analyzed on this and other blogs.
    It is somewhat similar to the MSM and their biases in criticizing (and criticism with which I would agree) some administrations and politicians while withholding it from others.

  11. Jeff

    Maybe I’m missing something, but I think you are being slightly unfair to McLean et al about the use of derivative series. They state in their original paper, and reiterate in their (unpublished) response to Foster et al, that the derivative series were only used for determining time lags, and that the high correlations they quote relate to the derivatives, not to the raw data. But I agree with you that McLean et al’s conclusions were very likely overstated.

    As far as I can make out, McLean et al based their conclusions about the longer term link between SOI and global temperatures on their Figure 7, which showed that the (inverted and scaled) SOI graph line reasonably matched that of the graph line for the Global Tropospheric Temperature Anomaly (GTTA) as to overall change over December 1979 to June 2008 as well as (save during volcanic events) to short term fluctuations. I am not sure whether you, understandably, thought that this graph also was based on derivative series?

    In fact, as the description below Figure 7 correctly indicated, the main results graphed in that figure do not use derivative series. Confusingly, the SOI line in figures 7(b) & 7(c) is labelled “SOI derivative” in the figure 7 from the copy of the original JGR paper at http://icecap.us/images/uploads/McLeanetalSPPIpaper2Z-March24.pdf but as plain “SOI” in the graph in their response at https://noconsensus.files.wordpress.com/2010/03/mcleanetalsppipaper2z-march241.pdf, which McLean et al claim there to be a reproduction of Figure 7 from the JGR paper.

    I have been unable to access the official version of the JGR paper to check which is right. But the substantial question is whether the original SOI data or the derivative thereof (or of the SOI 12 month moving average) was actually shown in figure 7. Having examined the original SOI data, I can confirm that McLean is correct in asserting that figure 7 did use the unprocessed SOI data (or something very close to it – certainly not any derivative).

    As shown in Mclean’s Figure 7, the MSU satellite global temperature anomaly was little changed (it actually fell slightly) over the December 1979 to June 2008 period that they took, whether one takes the middle or lower troposphere series (it is unclear which McLean used). Whilst its trend is considerably greater than that of the (inverted and scaled) SOI, they would have been much closer if the periods of divergence associated with volcanic episodes in the pre-1996 period were left out of the comparison. It is true that the post 1979 RATPAC-A GTTA data rose relative to the MSU satellite GTTA data, but the satellite data may well be the more accurate series. So, whilst Figure 7 doesn’t prove that there is any relation between long term changes in SOI and in GTTA, it appears at first glance not to show much evidence that GTTA is trending upwards relative to (inverted) SOI.

    Having said that, I think that various of Foster et al’s criticisms are valid. Certainly, the claim in the paper that taking the derivative removes noise makes no sense. Also, the paper is very light on detailed statistical evidence. And I don’t see that McLean et al validly establish that SOI “exercises a consistently dominant influence on mean global temperature”, although they present evidence that it may do so on shorter term changes thereto.

  12. Nic,

    Sitting in the delivery room, I believe the paper indicated that they used lower troposphere for everything. My only real problem with their paper was that the correlation and variance appears to be explained with the derivative data while there is a big difference in long term trend. If it weren’t for the strong conclusions on the explanation of a large portion of GW it would have been ok.

    It’s not like the rebuttal disproves the conclusion either. Tammie et team just didn’t like the result.

  13. Jeff

    Yeah, I agree with you.

    The paper indeed stated that they used lower troposphere MSU data, but the URL they gave as its source was for mid-troposphere (MT) MSU data!

    Incidentally, I think that the filter McLean used for determining lags and deriving correlations, by taking the difference over one year between 12 month moving averages, produced a second difference series of a sort (with a separation of about a year between each of the first, second and third data points involved).

    As for the quality of Foster et al’s rebuttal, I had a go at replicating their “Demonstration of the MFC09 Filter”, and the results for the unfiltered data were nothing like they reported. The correlation of raw SOI and noisy trending SOI was around 0.3, much higher than the 0.13 they quote (=R-squared 0.0171). And the correlation was 0.65 if a trend of 0.1 C/ decade, similar to the actual GTTA trend, was added instead of the very high 0.5 C/trend they added.

  14. Nic,

    That’s very nice and the kind of thing which needed to be done with the rebuttal. If the original authors points were closer to true than not and the rebuttal is allowed to stand, it would be a bad thing for global warming science.

    There really is only so much time in a day for those of us who don’t get paid for this kind of work. I need to check what you say, if the Foster rebuttal exaggerated the problems that much perhaps there needs to be a correction. If you’re in the mood someday. hint. I want to go after Mann09 still b/c it’s full of crap IMO, and am looking for a bit of time.

Leave a reply to Kenneth Fritsch Cancel reply