cc: Keith Briffa , Phil Jones , "Michael E. Mann" , Scott Rutherford date: Fri, 18 May 2001 09:58:38 -0400 from: Ed Cook subject: Re: Comments on "Extending NAO Reconstructions ..." to: Juerg Luterbacher Hi Juerg, My comments on the error limits was intended as much to elicit comments and discussion as anything else. So, I am not necessarily suggesting that you change anything in your paper. What you have done is fine as far as it goes. And as Mike says, it is a good first step. Again as I said previously, I may be completely wrong here, which I am willing to accept if someone can convince me that the defined histories of variation in the proxies used for reconstruction doesn't matter to the estimation and interpretation of the error limits. It is not obvious to me that that is the case. The existence of errors in the predictors also tells me that some other method of regression analysis might be tried which explicitly allows for errors in both the predictors and predictand. Doing this could strongly affect the SE of the estimate on the predictand in ways that I wouldn't care to predict. Much work needs to be done here. Cheers, Ed >Hello Ed > >I just came back and saw all your mails and the >interesting discussion on the uncertainty ranges >of the reconstructions. I will now see how to >proceed, I will calculate the 1sigma estimates >and then decide what to show. In addition, we >intend to show the interannual variability >rather than the low frequency variability, so >the errors are calculated for each winter. > >Thank you very much again for everything and till >later > >Have a nice weekend > >Juerg ================================== Dr. Edward R. Cook Doherty Senior Scholar Tree-Ring Laboratory Lamont-Doherty Earth Observatory Palisades, New York 10964 USA Email: drdendro@ldeo.columbia.edu Phone: 845-365-8618 Fax: 845-365-8152 ==================================