date: Tue, 07 Jun 2005 10:46:06 +0100 from: Tim Osborn subject: Fwd: Re: IPCC AR4 to: "Keith Briffa"

>Reply-To: "Rob Wilson" >From: "Rob Wilson" >To: "Tim Osborn" >Subject: Re: IPCC AR4 >Date: Tue, 7 Jun 2005 10:36:45 +0100 >X-Mailer: Microsoft Outlook Express 6.00.2900.2527 >X-OriginalArrivalTime: 07 Jun 2005 09:37:38.0421 (UTC) >FILETIME=[8B049A50:01C56B44] >X-Spam-Score: 0.1 >X-Spam-Level: / >X-Spam-Flag: NO > >Morning Tim, >Indeed, the data I sent you were the time-series generated from the simple >regression calibration >However, the data in Figure 3B are scaled time series. I mentioned it in >the caption of that figure. > >Essentially I used the regression based calibration to undertaken >verification and generate the error bars. >Although verification could be made using scaled series, I could not >generate error bars with scaling. > >So which version to use? I think the scaled one likely would better >portray a better estimate of temperature amplitude. >However, this is really only a gut feeling and has certainly not been tested. > >I have essentially used the same methodology for the coral paper and >Philip Brohan is generating a pseudo-proxy data-set with which I can test >regression vs. scaling procedures. This might become a part II paper, >depending on the results. > >re. my Figure 4 - I scaled all recons and models as I felt it was the >better method of comparison. I am not sure if others would agree, but as >the models show higher variance than the instrumental data and most recons >lower, I need to scale them all so I was comparing apples with apples with >regards to past temperature amplitude change. > >Hope this clarifies everything >Rob > >PS. between you and me, I don't think our new NH recon really adds >anything new expect perhaps the fact that we should not put too much >confidence in these recons prior to 1100 or so. After seeing Ulrich >Cubasch's talk on the 32 flavours of climate, I think the whole >methodology is up in the air as well. Do you know if he has drafted a >paper on this work yet? > >----- Original Message ----- >From: Tim Osborn >To: Rob Wilson >Sent: Tuesday, June 07, 2005 10:07 AM >Subject: Re: IPCC AR4 > >Hi Rob, > >I haven't been able to reproduce figure 3B in your paper then. This is >labelled as anomalies wrt 1961-1990 and thus if I plot the data you sent >for "RCS recon" "unfilter calib" I expected to reproduce it. In 3B there >are numberous positive RCS values in the 20th century, yet only a handful >in excel file you sent. > >I'm wondering whether this is because you calibrated the reconstructions >using regression against instrumental temperatures - and that is the data >you sent - but then you scaled to match the instrumental mean and variance >- and that is the data in 3B. Is this right? If so, I'd need to know your >view on which you would prefer to be presented in the IPCC figure - >presumably you prefer the scaled version, since that's what you show in the >figures of the paper. But you also scale all the others for comparison >(figure 4A), whereas we've decided with discussion from the other chapter >authors to stick with original calibrations. So please let me know if I'm >right about reasons why I can't reproduce your figure 3B and if so which >version you and Rosanne prefer to show. > >I'm also a little unclear why you would regress and then rescale, since the >latter would yield the same results as rescaling the original series >wouldn't it? Or was it a multiple regression? Sorry, I've only read your >paper once and didn't pick up on this yet! Still, this is more a >scientific issue and the question about the IPCC figure is the more important. > >Cheers > >Tim > >PS. Will get comments to you about the coral work very soon! > > > > > > >At 20:54 06/06/2005, you wrote: > >Hi Tim, > >they are anomalies relative to the 1961-1990 mean, although remember > >calibration was made over the 1856-1978 period. Therefore, the 1961-1990 > >mean of the reconstructed series will likely not be zero - in fact due to > >the divergence, the are slightly lower. > > > >regards > >Rob > > > > > >----- Original Message ----- > >From: <Tim>mailto:t.osborn@uea.ac.uk>Tim Osborn > >To: > <Rob>mailto:rjwilson_dendro@blueyonder.co.uk>Rob > Wilson > >Cc: > <K.briffa@uea.ac.uk>mailto:K.briffa@uea.ac.uk>K.briffa@uea.ac.uk > >Sent: Monday, June 06, 2005 3:43 PM > >Subject: Re: IPCC AR4 > > > >Hi Rob - thanks very much for the data. One question - I'm using the > >unfiltered RCS calibated data - what are they anomalies from? The > >1961-1990 mean, or the calibration period mean? Cheers, Tim > > > >At 08:32 04/06/2005, Rob Wilson wrote: > > >Hi Tim and Keith, > > >I have attached the two NH reconstructions we have developed (STD and RCS > > >versions). > > > > > >I am also sending the original paper to give you more information if need > > >be. Unfortunately, and not surprisingly I guess, we heard yesterday that > > >it was rejected from Nature. We are currently editing for submission to > > >another Journal next week - under debate as to which. > > > > > >2 sigma error bars are provided for both the unfiltered reconstruction and > > >the 20 yr spline version (i.e. see Figures 2a and 2b in the supplementary > > >material). > > > > > >Please Note: The smoothed error bars were generated only as a guide for > > >the confidence of the lower frequency signal - I re-did the calibration > > >using smoothed series for each nested model. The resulting smoothed > > >reconstruction therefore does not overlay perfectly upon the unfiltered > > >reconstruction. In all analyses in the paper, I used the unfiltered > > >reconstructions or low pass filtered versions of them. > > > > > >if you have any questions please feel free to e-mail or phone. > > >regards > > >Rob > > > > > > > > >----- Original Message ----- > > >From: > > > <<Rosanne>mailto:druidrd@ldeo.columbia.edu>Rosanne>mailto:druidrd@ldeo.columbia.edu>Rosanne>mailto:druidrd@ldeo.columbia.edu>Rosanne > > > D'Arrigo > > >To: > <<Tim>mailto:t.osborn@uea.ac.uk>Tim>mailto:t.osborn@uea.ac.uk>Tim>mailto:t.osborn@uea.ac.uk>Tim > Osborn > > >Cc: > > > <<rjwilson_dendro@blueyonder.co.uk>mailto:rjwilson_dendro@blueyonder.co.uk>rjwilson_dendro@blueyonder.co.uk>mailto:rjwilson_dendro@blueyonder.co.uk>rjwilson_dendro@blueyonder.co.uk>mailto:rjwilson_dendro@blueyonder.co.uk>rjwilson_dendro@blueyonder.co.uk > > >Sent: Friday, June 03, 2005 2:44 PM > > >Subject: Re: IPCC AR4 > > > > > >hi Tim, > > > > > >Greetings. its is fine with me to show figures from our nh paper, which > > >we are submitting shortly. I am ccing to Rob Wilson who can send you the > > >data for the figures. > > > > > >cheers, > > >Rosanne > > > > > > > > > > > > > > > >Dear Rosanne, > > > > > > > >As you probably know, Keith Briffa has the job of putting together > > > >the section on the climate of the last 2000 years for the next draft > > > >of the IPCC Paleoclimate chapter. I'm helping out with some > > > >comments and figures - in particular, I am preparing a diagram > > > >showing various TAR and post-TAR reconstructions of NH temperature. > > > > > > > >I contacted Ed Cook to get the updated calibration of the Esper et > > > >al. reconstruction, which he provided. But Ed also suggested that I > > > >contact you because you have a new tree-ring based NH temperature > > > >series in the pipeline. > > > > > > > >I'm not sure what stage it's at, but if it's appropriate would you > > > >be willing to have it included in an IPCC comparison diagram? To be > > > >included, a paper describing it would need to have been accepted for > > > >publication by December. > > > > > > > >Please let me know what you would like me to do: include it or not? > > > > > > > >If yes, then the data that I'd need are: > > > >(1) the calibrated NH temperature reconstruction (plus an indication > > > >of the reference period) > > > >(2) uncertainty ranges for the calibrated series, if you have them, > > > >at an appropriate time scale (e.g. for 30 year smoothed data) > > > > > > > >Best regards > > > > > > > >Tim > > > >Dr Timothy J Osborn > > > >Climatic Research Unit > > > >School of Environmental Sciences, University of East Anglia > > > >Norwich NR4 7TJ, UK > > > > > > > >e-mail: > > > <<t.osborn@uea.ac.uk>mailto:t.osborn@uea.ac.uk>t.osborn@uea.ac.uk>mailto:t.osborn@uea.ac.uk>t.osborn@uea.ac.uk>mailto:t.osborn@uea.ac.uk>t.osborn@uea.ac.uk > > > >phone: +44 1603 592089 > > > >fax: +44 1603 507784 > > > >web: > > > <<http://www.cru.uea.ac.uk/~timo/>http://www.cru.uea.ac.uk/~timo/>http://www.cru.uea.ac.uk/~timo/>http://www.cru.uea.ac.uk/~timo/>http://www.cru.uea.ac.uk/~timo/>http://www.cru.uea.ac.uk/~timo/>http://www.cru.uea.ac.uk/~timo/ > > > >sunclock: > > > > > > <<http://www.cru.uea.ac.uk/~timo/sunclock.htm>http://www.cru.uea.ac.uk/~timo/sunclock.htm>http://www.cru.uea.ac.uk/~timo/sunclock.htm>http://www.cru.uea.ac.uk/~timo/sunclock.htm>http://www.cru.uea.ac.uk/~timo/sunclock.htm>http://www.cru.uea.ac.uk/~timo/sunclock.htm>http://www.cru.uea.ac.uk/~timo/sunclock.htm > > > > > > > > > > > > >Dr Timothy J Osborn > >Climatic Research Unit > >School of Environmental Sciences, University of East Anglia > >Norwich NR4 7TJ, UK > > > >e-mail: > <t.osborn@uea.ac.uk>mailto:t.osborn@uea.ac.uk>t.osborn@uea.ac.uk > >phone: +44 1603 592089 > >fax: +44 1603 507784 > >web: > <http://www.cru.uea.ac.uk/~timo/>http://www.cru.uea.ac.uk/~timo/>http://www.cru.uea.ac.uk/~timo/ > >sunclock: > ><http://www.cru.uea.ac.uk/~t > imo/sunclock.htm>http://www.cru.uea.ac.uk/~timo/sunclock.htm>http://www.cru.uea.ac.uk/~timo/sunclock.htm > >Dr Timothy J Osborn >Climatic Research Unit >School of Environmental Sciences, University of East Anglia >Norwich NR4 7TJ, UK > >e-mail: t.osborn@uea.ac.uk >phone: +44 1603 592089 >fax: +44 1603 507784 >web: http://www.cru.uea.ac.uk/~timo/ >sunclock: >http://www.cru.uea.ac.uk/~timo/sunclock.htm Dr Timothy J Osborn Climatic Research Unit School of Environmental Sciences, University of East Anglia Norwich NR4 7TJ, UK e-mail: t.osborn@uea.ac.uk phone: +44 1603 592089 fax: +44 1603 507784 web: http://www.cru.uea.ac.uk/~timo/ sunclock: http://www.cru.uea.ac.uk/~timo/sunclock.htm