cc: Anders , Eduardo.Zorita@gkss.de, hegerl@duke.edu, esper@wsl.ch, k.briffa@uea.ac.uk, m.allen1@physics.ox.ac.uk, t.osborn@uea.ac.uk date: Mon, 07 Aug 2006 13:35:09 +0000 from: Nanne Weber subject: Re: Another draft -- hopefully ready for submission to: Martin Juckes Hi Martin, thanks a lot for the new draft. I have a number of smaller comments, as indicated in the Latex file (marked by %CNANNE....%CENDNANNE, so that you should be able to find them easily). Also some bigger comments: 1) I think that we should avoid entering into unresolved disputes (unless we really want to present something new as in section 4). Otherwise it might provoke endless comments and emails, similar to what Anders is referring to. Therefore, I propose to shorten the first para of section 2.8, just stating the two views (of Storch et al and Mann et al) and noting that the debate is ongoing (see text in Latex file). 2) I found the review of MMetc. in section 3 quite accurate, but agree with Anders that we should be careful here. As another example of what can be provoked I refer to the just submitted paper of Burger and Cubasch on the CPD website, which has already attracted a number of heated+lengthy+technical comments from 1 referee, plus counter-comments 3) table 1 in section 3 is simply confusing to me. I propose to a) order the data according to geographical location b) add references that are still missing, especially for the Hegerl et al data (paper not published yet) and Esper et al (original paper does not contain data soure references) c) explain in text that 2 versions of same treering record (Tornetraesk, Taymir, Urals) mostly differ in treatment of long timescales Example reshuffled table is attached 4) section 4 is really nice, concise and clear in its goals. However, I got lost in the statistics. I have put a lot of dumb questions in the text where I got lost. Hope that you can clarify these. After all, we are trying to reach a general readership. 5) Figures: could you do 6+7 with the same vertical scale? I had to look hard for the bars in Fig. 6, maybe replace them with something more solid (filled dots?). The graphs would be better visible if you would place the coloring code in the caption instead of in the figure. Fig. 8: it seems more relevant to me to give the autocorrelation for lags of 0-30 yrs (about), as these can be reliably estimated and just discuss some of the features for longer lags in the text (maybe even the Appendix). Best, Nanne =================================================== Attachment Converted: "c:\eudora\attach\table1_mitrie.pdf" Attachment Converted: "c:\eudora\attach\mitrie1.tex"