13 November 2009

McKitrick on Amplification Ratios

Ross McKitrick has added his two cents on the criticisms raised by Gavin Schmidt at Real Climate, and they show what a tempest in a teapot this all is (bold added):
AMPLIFICATION RATIOS: I was drawn into the dispute between Gavin Schmidt and Klotzbach et al. (see Pielke Sr.) over the latter paper's conclusion that the surface temperature record over land has a warm bias for the purpose of measuring global warming. I was cited in Klotzbach et al. as the source of a claim that the GISS model exhibits amplification over land of about 1.2. I should not have been cited, since all I did was report in an email to John Christy the average trop/surf trend in Gavin Schmidt's own GISS data pertaining to my 2007 surface temperature analysis. The information source, in other words, was Schmidt himself, not me; and in any case I did not provide it as a personal communication for the purpose of a journal article (which I did not know was being written). Phil Klotzbach and his coauthors have issued a correction on this point. In subsequent correspondence with Gavin Schmidt he reported to me that he had corrected an error his original IJOC archive and also that the GISS model classifies land differently than CRU so some of the 440 grid cells are actually over ocean in his model. He supplied me with the GISS landmask. I have recomputed the original results using the corrected data and the GISS landmask. The cosine-weighted amplification ratio over land is about 1.106 and over ocean is 1.602, where 'land' and 'ocean' are according to the GISS landmask applied to the 440 grid cells used in my 2007 paper.
So in the paper we used an amplification of 1.2 based on calculations done earlier this year as described above. Subsequently, Gavin computed a value that was apples to our oranges and arrived at 0.95. Ross McKitrick just now re-computes a value that is more apples-to-apples and arrives at 1.1. We have shown that our conclusions are insensitive to the choice of 1.2 and 0.95 and you can easily conclude that using 1.1 instead won't matter either to the conclusions.

It is virtually certain that Gavin will contest the new number from McKitrick. Does it matter to me which one is judged to be most appropriate? From the standpoint of the analysis of Klotzbach et al., no. This sort of dispute won't be resolved on blogs, but in the peer reviewed literature where it belongs.

In other news, we've been informed by Ben Santer that our paper contains a referencing error (i.e., Santer et al. 2005 is used in the introduction where it should instead be Santer et al. 2000) and a typo (i.e., it reports a "p-test" rather than a "t-test"). We regret these errors, but certainly appreciate the close reading. We hope that the substance of the paper receives similar close attention. Once again, I hope that Gavin submits his critique of our paper's analysis and conclusions to the scientific literature, and if his work improves upon our work leading to better understandings, then good for science.


  1. Roger-I'm not sure if you knew this already but it's interesting that Ross and Gavin are in communication. On an issue which is, in fact, related to your paper, they have been having an interesting disagreement in the peer reviewed literature. In 2007, Ross and Pat Michaels published a paper showing that socio-economic signals caused substantial artificial warming. Gavin in 2009 published a response which alleged (using arguments which are in fact erroneous, as I will get to later) that the conclusions were spurious. Now Nicholas Nierenberg and Ross have a submission to IJOC which shows specifically where Gavin's paper is in error.

    The papers are:

    McKitrick, Ross R. and Patrick J. Michaels (2007) Quantifying the influence of anthropogenic surface processes and inhomogeneities on gridded global climate data.Journal of Geophysical Research-Atmospheres, 112, D24S09, doi:10.1029/2007JD008465.

    Schmidt, Gavin (2009) Spurious correlation between recent warming and indices of local economic
    activity. International Journal of Climatology 10.1002/joc.1831

    McKitrick, Ross R. and Nicolas Nierenberg (2009). Correlations between Surface Temperature Trends and Socioeconomic Activity: Toward a Causal Interpretation . Submitted to International Journal of Climatology.

    Anyway, what I find interesting is that they actually can engage in exchanges which, apparently, are relatively civil. Where exactly is that Gavin? Perhaps his attitude is trying to keep face at RC, where he has a reputation to uphold as Fidei Defensor?

  2. Andrew:
    First I would say that most disagreements in journal articles appear relatively civil and tend to be less pointed than in one-on-one interactions whether on blogs, panel discussions or in workshops.
    Your point about Gavin's role as Fidei Defensor is IMHO spot on. It is naive to believe that scientists are somehow immune to or can compartmentalize the motivations and normative biases that afflict the rest of us. Gavin's bias is simply more visible when it comes to discussing anything that is close to policy decisions. Plus Gavin may simply have an exceedingly thin skin.
    Finally, based on his performance in the debate he had with Crichton, Lindzen and Stott, I am not sure that Gavin is quite as facile in the wit department as he would like to be. The turnaround in voting on that evening has to be seen to be believed. One explanation for the change in voting from against the motion that Global Warming is Not a Crisis is that the audience - like a jury - did not like Gavin's demeanor. He was in the debate incredibly condescending, used silly metaphors and analogies, accused his opponents of essentially being liars or tools of special interests and dripped sarcasm - which we all know is the lowest form of wit. (The debate can be found on You Tube.)

  3. Roger,

    McKitrick's numbers seem sort of strange to me in that, once you use the fact that the earth is about 70% ocean and 30% land, you would get a global average amplification of ~1.45 based on his land and ocean values. Doesn't this sound a little high?

  4. -3-Joel

    You'd have to ask Ross. Do note that he uses a cosine weighting and a subset of land surface as represented by the overlap of CRU observed and the GISS model. I am under the impression that everyone agrees that the global value is about 1.2.