Comments on: Jaemtland, an Esper Site

By: John G. Bell

John G. Bell — Thu, 15 Sep 2005 01:02:43 +0000

I like your grass plots. For some time I’ve wondered about CO2 fertilization at some of the sites. You have addressed this several times and well. Perhaps some graphic might be made that demonstrates or suggests CO2 fertilization. My thought is to plot a site’s data in 3D. X – RW, Y – MXD, Z – Year. Color data points scaled to CO2 in that year. A sort of 3D colored grass diagram. You may have to plot the average around a year for a tree’s RW or MXD so that the individual trees can be followed in the diagram. A twisted idea? Fun to rotate it around and view it.

By: Steve McIntyre

Steve McIntyre — Wed, 14 Sep 2005 11:19:08 +0000

I’ve been attempting some more to decode the Science 2002 article: the linear/nonlinear has to do with the ring sidth (RW) series; there’s no apparent use of MXD in Esper et al 2002. If you do plots of individual cores – see for example my plots of individual cores for Gaspe, you get a wide variety of results. I have no idea how one would go about trying to divide these series into "linear" and "nonlinear" or even what purpose is serve by this distinction. I’m having difficulty seeing what the active ingredients are for Esper’s 20th century being higher than his 11th century – his max occurs in 980 (which is conveniently excluded from the millennial discussions). I’ve done plots for 5 sites now without seeing any with loud 20th centuries. There are 2 California foxtail (bristlecone cousin) pine sites by Graunlich, which are unarchived. I tried to get info from Graumlich a long time ago about sites that were sampled about 15 years ago and she said that she was still planning further publications and would not disclose the data. Esper also uses the Polar Urals data set and Jacoby’s Mongolia data set so there are a cople of the standard overlapping proxies. I can’t think of any reason for creating a class distinction between linear and nonlinear, but I’ll bet it affects the relationship between MWP and modern levels somehow.

By: Douglas Hoyt

Douglas Hoyt — Wed, 14 Sep 2005 10:54:05 +0000

If you plot MXD versus 1/RW^0.333, do you get a constant? If so, perhaps then it is considered linear, and if not, perhaps it is nonlinear?

By: TCO

TCO — Wed, 14 Sep 2005 02:00:05 +0000

(I’m sure this is already done in basic dendro work) but would be interesting to look at the visual issues with samples (or trees) that correspond to different behaviors. For instance are the alpha trees in certain locations (unshaded). anything else? Sometimes when one is troubleshooting something in the plant, and one gathers all these interesting statistical inferences…if you then feed it back to a visual inspection/walktrhough…you get some insights.

By: John A

John A — Tue, 13 Sep 2005 23:54:40 +0000

I suppose “linear” could mean “I think the relationship between temperature could be seen by eye” and non-linear the reverse.

With such a large amount of statistical noise, can anyone discern what relationship can there be with temperature or precipitation?

Also it appears that if “alpha trees” predominate the proxies after out-competing most others, could the extra fertilization of nutrients and carbon dioxide from the breakdown of the others cause a non-climatic boost to the growth of the alpha trees?