cc: k.briffa@uea date: Fri Apr 23 11:28:54 1999 from: Tim Osborn subject: Your Perspective for Science to: juppenbrink@science-int.co.uk, bgendlem@aaas.org Dear Brent and Julia, Keith is away today, so I am replying on our behalf. I have, however, spoken to Keith on the 'phone, to pass on your messages and to formulate our reply. >>Date: Thu, 22 Apr 1999 15:51:29 -0400 >>From: bgendlem@aaas.org (BGENDLEM) >> >> Julia has writtent the following abstract that will only appear on the >> web and she wanted your comments to make sure it was accurate. The abstract is excellent and no alterations are necessary. >> She also said she wanted you to address some point about expanding the >> caption to state which records are annual, seasonal, extra-tropical, >> etc. Could you let me know your response? >From: "Julia Uppenbrink" >Date: Thu, 22 Apr 1999 17:45:22 +0100 > >One point remains. Could you address JU15(2), which asks for a clarification >in the caption, i.e. which records are seasonal, annaul, extra-tropical... ? >both Raymond and Mann thought that this was important and I agree. Could you >send the necessary changes to Brent Gendleman, as I will be out; it will >likely be just a word here or there. If each reconstruction represents a different area and/or season, then intercomparison will show a combination of (i) real differences because the climate of each season and/or area really has varied in a different way, and (ii) differences due to inaccuracies in the proxy data used. For the purposes only of this piece in Science, where intercomparison of the records is the reason for showing the figure, we have attempted to reduce item (i) by re-calibrating each record against a common season and area and using a common period of data. This is all stated at the start of the caption, and it does succeed in reducing the differences between the various records. We can justify this approach, purely for the purposes of intercomparison, on the basis that the season and area that each record represents is somewhat arbitrary because they combine proxy records with different seasonal responses and because none of them have complete spatial coverage. The season/area is less arbitrary for the Mann et al. record because of their method of calibration, but we justify our re-calibration of their record by that fact that the table shows that it is still the best record in terms of its correlation with instrumental temperatures. This detailed justification is too long to go in the article, but I have provided it here so you can decide whether the caption needs to be altered. As a result of our re-calibration, one could argue that they are all estimates of the Apr-Sep temperatures averaged over land north of 20N, as we state. Nevertheless, the proxy records on which they are based do have varying spatial coverage, as we already state: "...circum-Arctic...northern NH...NH temperature proxies...global climate proxies...northern Eurasian...". We could expand the caption to state the seasons they represent, but all except the Mann et al. record represent warm season temperatures. Our suggestion would be, therefore, that only the exception needs to be stated. The caption could then become: "...are in red, global climate proxies [1000-1980, originally calibrated against annual temperatures, from (5,6)] are in purple, and an..." >> Also, the piece appears to have grown larger than space will allow, >> so, on Monday, we may need you to make some cuts. >If we have to shorten we will do so very carefully. As Keith noted, we couldn't see a whole paragraph that could simply be cut, and so we would be keen (of course!) for you to find room for it all. If that isn't possible, then we will have to go through it on a sentence by sentence basis trying to save words where we can. Please make any suggestions if you can identify parts that can be shortened. >We will fax the close-to-final version to you as I will be out and cannot >check it myself. We would appreciate seeing that; also, we've not yet seen the table. Best regards Tim