Plots of Gaspe and Polar Urals Data

Just as an experiment, I plotted up all the ring widths for the Gaspé and Polar Urals sites, with each core displaced a little. What I was thinking of the type of graph that you see in seismic surveys. The look is interesting, given how much weight is placed downstream on these data sets in multiproxy reconstruction.

First, the Gaspé data set which has a very strong hockey stick shape. Quite a few series do have strong 20th century growth, but this is far from general. Quite a few don’t. Prof Larson, a renowned cedar expert at the University of Guelph, said that cedars like cool and moist climate, rather than hot summers. In the cedars that he’s studied, he said that they grew slowly when they were young and when they were old. An issue that I’m pondering with many tree ring sites is the effect of standardizing by fitting a negatively sloped curve and the end effect bias if the slope is too negative. You can see some examples here, where early-period growth, which might be attributed to early period warmth, would be massaged out by curve fitting and attributed to being a juvenile growth effect.

Figure 1. Gaspé (cana036)

The next one is the Polar Urals site. Here I haven’t figured out yet how to make a longer scroll (I haven’t tried hard) so that the individual series are not squished so much. However, you can readily see the shortness of the individual cores at the Polar Urals site and that the early portion is poorly represented. This series has a relatiely strong inter-series correlation in the well-dated later section.


  1. TCO
    Posted Sep 11, 2005 at 11:24 AM | Permalink

    I often don’t get the implication of what you are saying. Short cores. interspecies corrleation. Is that good or bad?

    Also, don’t they normalize the proxy studies for tree age? Seems relevant…

  2. Steve McIntyre
    Posted Sep 11, 2005 at 11:30 AM | Permalink

    This refers back to crossdating. See my post on Polar Urals – crossdating.

  3. TCO
    Posted Sep 11, 2005 at 12:09 PM | Permalink

    saying the noun twice does not make it penetrate my thick head more. I need some verbs and sentences to make it mean something…

  4. Steve McIntyre
    Posted Sep 11, 2005 at 12:38 PM | Permalink

    If all your cores are 500 years long and you’re given a new core that’s 500 years long, you can crossdate it very accurately by wigglematching (i.e. correlations). But when your cores are on average

  5. TCO
    Posted Sep 11, 2005 at 12:45 PM | Permalink

    So you are questioning (or looking at) the quality of the dendrochronology itself?

  6. DAV
    Posted Sep 11, 2005 at 2:49 PM | Permalink

    But when your cores are on average …


    Was there supposed to be more, Steve?

  7. Posted Sep 11, 2005 at 2:51 PM | Permalink

    re 4: Looks like your sentence was truncated..

  8. Posted Sep 11, 2005 at 2:52 PM | Permalink

    So I am not the only one who noticed that 😉

%d bloggers like this: