For anyone interested in historical temperature measurements the
following paper is a must read:
Because the very obvious different between temperature estimates based
on tree ring proxies and other proxies appear to throw into doubt any
historical temperature scale based on tree-ring proxies.
To put it in context, most tree-ring estimates of historical temperature
have a shape which barring the odd bit of noise appears to be completely
stable until the 20th century after which there is a sharp rise
(coinciding with the period of increased CO2 which is known to help
In contrast this temperature estimate using derived solely from proxies
which did not mainly use tree-rings shows the kind of temperature
changes historians recognise from accounts of the medieval warm period
and the little ice-age.
In explaining the difference the author suggests that whilst Trees do
respond to temperature and so tree rings do change with changes in
temperature, over a long period of time the forest itself will
self-adapt so that if periods encourage greater tree ring growth, more
trees will grow until the number reduces the rate of tree growth to the
long-term optimum for the species. Similarly, if climate changes to
reduce tree growth, forest's thin out so that individual trees then have
more chance of growing.
So the number of trees adapt to climate so as to tend to allow trees
always to grow at their optimum rate. SO ONLY IN PERIODS OF CHANGING
CLIMATE WHERE FOREST DENSITY HAS NOT HAD TIME TO ADAPT DO TREE RINGS
REFLECT ACTUAL TEMPERATURE CHANGE.
Obviously at the very least this throws into doubt any historical
temperature scale based on tree-ring proxies.