from: Keith Briffa <k.briffaatXYZxyz.ac.uk>
subject: Re: Proxy data question
it is , as you imply, only "useful" to look at proxies before the availability of climate
data - as they are meant to give some information on climate when we have no other - ie
they are used as palaeoclimatic substitutes. Yes , one needs to compare them with the
"target" climate to be able to judge how well or poorly this substitution is . The formal
regression procedures involved in "calibrating" and "verifying the calibration results"
provide insight into how well the inferences on past climate are likely to stand up (as
shown by quantified uncertainty bounds on our reconstructions ) - provided of course that
the "uniformitarian principle is maintained. The reason Mann only used the proxies up
until about 1980 , is because they did not extend through to the present day at many sites
, and 1980 was a convenient cut off to use in his calibration . So we can not say how well
the proxies would mimic the recent (post 1980) warming. This is the point I stressed in my
lecture - that we need to update many crucial proxies , and then test the relationships we
have derived for retrodicting climate - by using them to estimate the warming in different
areas (and the globe) in the last 20 years and comparing the estimates with measured
reality.This is likely to be a harsh test and will likely show that we may underestimate
the true magnitude of the warming - but whether to an extent that exceeds the calculated
uncertainty in the past regression estimates is a moot point.
At 17:37 24/10/2005, you wrote:
I hvae a question related to your lecture and the M525 coursework which I
wondered if you could help me with.
How recently are proxy indicators (specifically tree rings) used as a
record of climate? What I mean is, can proxy records be used for the last
20 years? I realise there is instrumental data for this period but is it
possible to use proxy data to look at temperature in this most recent
period as well? Does it match with the instrumental record for this
period? I think I read in Mann et al 1998 that it was only used up to 1980
- what is the reason for this, is the time lag between change and response
Hope that makes sense!
Thanks for your help,
Professor Keith Briffa,
Climatic Research Unit
University of East Anglia
Norwich, NR4 7TJ, U.K.