date: Wed, 04 Jun 2008 10:50:37 +0200 from: Kurt Nicolussi subject: Re: Divergence to: Tom Melvin Hi Tom, thank your for your mail. The fine weather (sunny and warm) is over at the moment, but I managed to go into field yesterday before it started raining. > I criticised Walter Oberhuber's methods during his talk. He found > divergence using curve-fitting methods. He gave me a copy of his paper > (2008 - Trees). He has a good set of trees, suitable for demonstrating > the absence of divergence. Do you work with/know his group? Tomy told me about your discussion with Walter - of course, I know him a little bit - in the early 90ies, I personally worked about 1 year at the Institute of Botany, than I went back to a better job at the former Institute of High Mountain Research, and Walter got be former job at the Botany. He is more focused on physiological questions and recent time periods - at least up to now. > > Also I tried a transform with simple RCS method. The latest 140 years of > tree growth do not fit the lower growth levels of the previous > millennia. See attached. I will try other standardisation e.g. multiple > curves etc and also try and find lower climate values to produce a wider > distribution. I don't believe in simple RCS, as you know. one wish: - could you write a procedure for cruRCS, that it is possible to choose the splitting of series for multiple RCS (e.g. three RCS-curves: within +/- 1 standard deviations is one group, and the two other groups are outside of these limits) Best regards Kurt