From: Malcolm Hughes To: Tom Melvin Subject: Re: recent paper Date: Mon, 21 Sep 2009 15:23:49 -0700 Cc: Keith Hi Tom - please find the Esper article in question attached. The so-called Indigirka River data set is not yet available because it has not been published. I am currently working on that with Russian colleagues, and was indeed in Switzerland the week before last to work with one of them on specifically this. All being well, there will be an accepted manuscript before next summer, and at that point I will make the data freely available. Once we get to that point, I'll let you know, of course. Cheers, Malcolm Tom Melvin wrote: > Malcolm, > > 1. There was a recent Esper Siberian paper I recall reading but I > cannot find it at the moment (my comment was on the Divergence > pitfalls paper). I will find the paper and see if there is an > explanation. > > 2. For trend distortion to produce a "divergence" effect there needs > to be a distinct increase (or decrease) over the last few decades of > growth, e.g. at TTHH and curve fitting methods should be used. In the > attached figure the Scandinavian site groups (red) have an increase at > 1920 and are likely to show divergence using curve fitting methods. > Some of the eastern most chronologies might also show divergence if > 250+ year old trees were used. > > 3. RCS should not produce "divergence" over decades as an artifact if > sub-fossil trees are used. RCS on modern chronologies has all sorts > of bias. We have lots of ideas to test in the divergence project and > lots of data to test them on. > > 4. Keith has been complained at by Climate Audit for cherry picking > and not using your long Indigirka River data set. Not used because we > did not have the data. Please, could we have the data? We will make > proper aknowledgement/coauthorship if we use the data. > > Tom > > > > At 16:35 21/09/2009, you wrote: >> Tom, I don't disagree with your take on the lack of originality of >> much of what is in the paper. The question is: why is there apparently >> divergence in ring width in some of this region in Briffa et al 98 but >> not in this paper? Isn't espers failure to see divergence >> counterintuitive when using RCS in his way? >> Cheers, Malcolm >> >> >> On Sep 21, 2009, at 2:11 AM, Tom Melvin wrote: >> >>> Malcolm, >>> >>> The Esper "Divergence pitfalls .." paper does not appear to add >>> anything of significance. None of the figures show any form of the >>> divergence discussed in papers e.g. a recent (last few decades) >>> change in the slope of tree-ring growth indices compared to climate. >>> Differences in overall slope, generally weak relationships, >>> differences in variance, and the effects of using selected >>> calibration periods are all problems to be addressed in >>> reconstructions but are not divergence. >>> >>> I cannot foresee needing to reference this paper in discussions of >>> divergence as all the suggestions have more detailed, earlier >>> references. >>> >>> Tom >>> >>> >>> At 22:33 18/09/2009, you wrote: >>>> Hi Tom - I had a good talk with Keith on the phone the other day, >>>> mainly to wish him well. He did suggest I ask you for your take on >>>> the recent Esper et al paper on divergence (or rather the lack of >>>> it) in Siberia. Looks like the problem disappears. WHat do you >>>> think? Cheers, Malcolm >>> >>> Dr. Tom Melvin >>> Climatic Research Unit >>> University of East Anglia >>> Norwich, NR4 7TJ, U.K. >>> >>> Phone: +44-1603-593161 >>> Fax: +44-1603-507784 > Dr. Tom Melvin > Climatic Research Unit > University of East Anglia > Norwich, NR4 7TJ, U.K. > > Phone: +44-1603-593161 > Fax: +44-1603-507784 Attachment Converted: "c:\eudora\attach\Esper-2009-GCB.pdf"