Detrending of ring width curves
The innermost ring widths of a tree are usually also the widest ring widths of that tree.
Detrending of a ring width curve is the process of modifying the curve to remove the effects of tree aging
from the ring width data. This detrending also includes a scaling of the ring width curve to compensate for the
effects of rich or poor soil.
A detrended curve contains information about climate history, a matter which is in focus of today's research.
Though not only climate history is comprised in a detrended ring width curve!
If many trees are removed from a forest then the resulting increase in growth for those trees left
can easily be mistaken as being the result of a sequence of warm and rainy summers.
If many trees are attacked by insects or fungi the growth rate goes down.
This can later be incorrectly understood as the result of a sequence of very cold summers.
You have good reasons to be suspicious when you interpret a ring width curve. There may
be unexpected causes behind a change!
You can use the new detrending mechanims in CDendro to get a grasp of how your trees have grown. The detrended ring width curves may help you with crossdating.
In principle, the detrending of a ring width curve is done by dividing each ring width value with the
mean width ring value of the whole tree.
This calculation makes growth curves from different trees
comparable though their mean ring width values differ greatly because of different growth conditions.
This will compensate for the decrease in growth related to aging.
This divides each ring width value with the mean value of the nearest surrounding 10 ring widths.
At the very ends of the ring width curve, division is done with the mean value of the nearest 5 ring widths.
If you like, you can save any detrended ring width curve as a .wid-file
Note: When one of the detrend options is turned on for a sample which itself is the reference then
the shape of this reference ring width curve will be plotted the same way in all other sample windows! So if you
want to compare Neg-Exp-deTrended ring width curves for a sample and a reference, you have to see that
both sample and reference have their "NExp detrended curve" options selected.
Of course, if the reference curve is based on e.g. a .d12 file containing an already Neg-Exp:ed ring width curve it should
be displayed with the "Original ring widths" option.
(After a change of option in another window, you may have to click the "Refresh curve" button when you are back at the curve you are analyzing.)
Collections:
You do want detrending when you take a mean value curve from many different trunks.
You do not want detrending when you analyze several radii from the same trunk.
When possible you will probably measure several radii of a stem. When you add such ring width series to
other series based on a single radii per stem the multi-radii-stems will be overrepresented in your sum.
To avoid this you can turn on the "Sum by stem" option before you click "Create a sample
with width mean values and normalized data".
CDendro will then first make a per-stem mean value calculation
(of normalized and of scaled ring width values) and then add these per-stem mean values to the definite result.
CDendro assumes that sample names like NMVK04A, NMVK04B, NMVK04C come from the same stem. I.e. the radius identity is represented by a letter after a digit. Even NMVK04D1 and NMVK04D2 are considered radii of the same stem.
In actual cases I have seen no significant difference between mean values from the collection as a whole or from a by-stem-calculation. What matters is correctly measured radii.