date: Mon Nov 3 18:28:04 1997 from: Keith Briffa subject: Re: to: Tom Wigley Tom thanks for the info. Actually this is a chance for me to to mention that we have for the last few months at least, been reworking the idea of looking in the Schweingruber network data for evidence of increasing tree growth and hence ,potentially at least, evidence of changing tree(read biomass) uptake of carbon. The results are dramatic - not to say earth shattering because they demonstrate major time-dependent changes - but changes that are consistent in different areas of the network. We have regionalised over 350 site collections , each with ring width and density data , age-banded the data so that we look only at relative growth in similar ages of trees through time and recombined the standardisd curves to produce growth changes in each region. Basically growth is roughly constant (except for relatively small climate variablity forcing) from 1700 to about 1850. It then increases linearly by about up until about 1950 after which time young ( up to 50 year old) basal area explodes but older trees remain constant . The implication is a major increase in carbon uptake before the mid 20th century - temperatue no doubt partly to blame but much more likely to be nitrate/Co2 . Equally important though is the levelling off of carbon uptake in the later 20th century. This levelling is coincident with the start of a density decline - we have a paper coming out in Nature documenting the decline . In relative terms (i.e. by comparison with increasing summer temperatures) the decline is represented in the ring width and basal area data as a levelling off in the long-timescale inrease ( which you only see when you process the data as we have). The density data do not show the increase over and above what you expect from temperature forcing. I have been agonising for months that these results are not some statistical artifact of the analysis method but we can't see how. For just two species (spruce in the western U.S. Great Basin area and larch in eastern Siberia) we can push the method far enough to get an indication of much longer term growth changes ( from about 1400) and the results confirm a late 20th century apparent fertilization! The method requires standardizing (localized mean subtraction and standard deviation division) by species/age band so we reconstruct relative (e.g. per cent change) only . We have experimented with integrating the different signals in basal area and density(after extracting intra ring ring width and density data where available) within a 'flat mass' measure which shows a general late 20th century increase - but whether this incorporates a defensible relative waiting on the different components (and what the relative carbon components are) is debatable. We now need to make some horrible simplistic assumptions about absolute carbon in these (relatively small) components of the total biomass carbon pool and imlpications for terrestrial and total carbon fluxes over the last few hundred years - and beyond! Without these implications we will have difficulty convincing Nature that this work is mega important. There are problems with explaining and interpreting these data but they are by far the best produced for assessing large scale carbon-cycle-relevant vegetation changes - at least as regards well-dated continous trends. I will send you a couple of Figures ( a tiny sample of the literally hundreds we have) which illustrate some of this. I would appreciate your reaction. Obviously this stuff is very hush hush till I get a couple of papers written up on this. We are looking at a moisture sensive network of data at the moment to see if any similar results are produced when non-temperature-sensitive data are used. You would expect perhaps a greater effect in such data if Co2 acts on the water use efficiency . At 09:30 AM 11/3/97 -0700, you wrote: >Dear Keith, > >Look at Tremblay et al. GRL 24, 2027-30 (1997) and Dyke et al. Arctic 50, >1-16 (1997). These papers deal with driftwood in the Arctic over the past >9000 years. They note that genera can be distinguished, but not species >Hence, they can't say where the wood comes from, North America versus >Europe. Surely cross-dating could do this? May be worth getting in touch >with Dyke et al. > >Tom > > > ********************************************************** > *Tom M.L. Wigley * > *Senior Scientist * > *National Center for Atmospheric Research * > *P.O. Box 3000 * > *Boulder, CO 80307-3000 * > *USA * > *Phone: 303-497-2690 * > *Fax: 303-497-2699 * > *E-mail: wigley@ucar.edu * > ********************************************************** > >