date: Thu May 20 13:56:35 2004 from: Phil Jones subject: Re: [Fwd: Re: Paper on ERA-40/NCEP/CRU comparisons] to: Adrian Simmons Adrian, Methinks they try to defend their paper too much ! The comment didn't say they are wrong, just not justified. Their argument about assimilation of the surface temperature data doesn't work either. Your report shows that changes in input data (surface and satellite and sondes) affects the fields. So, how do they prove that their differences are due to urbanization and land-use changes and not data inputs. The differences between CRU and ERA-40 can hardly be said to be gradual. You've explained most of them by data input changes. The jumps with NCEP are larger and there are more of them. I'm here today and tomorrow then off next week. Back for the whole of June and July (except for a fortnight when the house move occurs). If you want to give me a call after June 1 then do. How soon can the report go to press? Also, do you intend to go to Geneva for GIP in August? Cheers Phil At 23:04 18/05/2004 +0100, you wrote: Ming, Eugenia I'm travelling also, and do not have a copy of your Nature paper with me. I have looked at Fig 16 however. It does show a rise in the NH in the difference between temperature at two metres and near the top of the boundary layer. But the rise occurs principally rather sharply in the first half of the 1990s, not what I would have expected from an affect of change in land-surface use, and it is seen also in the simulation, which did not have any change over time in prescribed land characteristics. Unless there is simply chance agreement between the simulation and the reanalysis I suspect that there is some effect from the SST/sea-ice analysis at play here. The plot corresponding to Fig 16 but restricted to North America is attached. Best regards Adrian cai wrote: Dear Adrian, Eugenia is on travel. I am not sure if she would be able to check her email in time to response your email. Anyhow, this is my response to two additional questions of yours, which follows after your original questions below. Regards. Ming Cai On May 17, 2004, at 5:46 PM, Adrian Simmons wrote: Dear Ming and Eugenia Thanks for your comments, which I'll take note of when I revise the text of our paper in a week or two. I am puzzled, though, by your second comment. I do not understand the reference to Fig 6a, which does not relate to the trend above the boundary layer. The relevant figure is Fig12b,which shows a very similar trend over North America at the surface and at level 49. The trend is a little lower at 500hPa (attached), but not much. I'd appreciate it if you could clarify this point. We meant Fig. 16 instead of Fig. 6 in our comment letter. Looking at Fig. 12b or the figure you attached in your last email, I have to agree that the difference between ERA40 2 meter temperature and level 49 temperature indeed is small (I understand that the former is subject to the second tier assimilation and the latter results from the primary data assimilation system. This is indeed very impressive result). However, if we look at the red curve in Fig. 16a, we do see the difference between the 2-meter temperature and level 49 temperature increases in time over NH. From that plot, one can conclude that the 2-meter temperature warms at a fast rate than that at the level 49. Perhaps, we didn't interpret that figure correctly. I appreciate if you could help us to interpret that figure. Also, with regard to your fourth point, strictly speaking I did not write that our AMIP run argued against an effect of land-use change; I actually wrote that it did not support your conclusions. That was intended as a more netral comment, as one might indeed have reservations as to the conclusions one can draw from imperfect model simulations, just as one might have reservations about what can be deduced from imperfect reanalyses. Yes, we understand that that result is neutral in the context of our Nature paper is concerned. However, if you put that result in the context of our Nature paper, particularly saying that it didn't not support our conclusion, that would make the result NOT "neutral", which is a concern of ours. Best regards Adrian cai wrote: Dear Adrian, Attached is our response to your manuscript. Thanks again for sharing the manuscript with us. We are looking forward to more discussion on this subject. Regards. Ming Cai On Apr 25, 2004, at 3:36 PM, Adrian Simmons wrote: Dear Ming Thanks for the response. Unfortunately I shall not be at the EGS meeting. I myself will be travelling from 6 to 21 May, but hope to check my email from time to time while away. I will not take our manuscript any further until I have heard from you and Eugenia, but would like at least to publish it as an ERA-40 report (with or without reference to your work) as soon as I get back. Best regards Adrian Ming Cai wrote: Dear Adrian, Thanks for sending the manuscript to us. Since I will be in EGS meeting next week, Eugenia and I would not be able to get together to discuss your paper till my return from the trip. Meantime, I assume that you may be also attending the EGS meeting at NICE. If so, we could have a lunch or a brief discuss during the break. I will present a paper co-authored with Eugenia at 9:00AM on Thursday (April 29) morning in Session CL11 (Lecture Room Euterpe). The title is "Can reanalysis have anthropogenic climate trends without model forcing?" I think your work strengthens our arguments. I look forward to meeting with you and discussing with you on implication of your work on our Nature paper and on the work I am presenting at the EGS meeting. Regards. Ming Cai by "Cai, M; Kalnay, E" has been scheduled for an oral presentation in Session CL11, Lecture Room Euterpe on Thursday, 29 April 2004, 9:00. On Apr 23, 2004, at 5:53 PM, Eugenia Kalnay wrote: -------- Original Message -------- Subject: Re: Paper on ERA-40/NCEP/CRU comparisons Date: Fri, 23 Apr 2004 22:30:07 +0100 From: Adrian Simmons Organization: E.C.M.W.F To: Eugenia Kalnay References: <40894D36.2030005@ecmwf.int> <40897810.90705@atmos.umd.edu> Thanks for the initial response, Eugenia. I won't reply fully now, but will wait for your detailed comments. One thing though. I did note you replied to the comments in the references, but certainly could flag this in the text also. But let's see whether the references to your Nature letter survive at all. Maybe you'll convince me to say nothing. And a second. I show in the paper that the trace of temperatures over North America at model level 49 (about 850hPa for 1000hPa surface pressure) are virtually the same as at the surface. The level-49 analysis is hardly affected at all by the fact we did a separate analysis of two-metre temperature using SYNOPS. There's just a weak affect at this level from the soil-moisture and soil-temperature analysis. Am I wrong to expect that land-use changes would affect near-surface temperatures but have less impact on temperatures at the top of the PBL? There's no chance of the paper being submitted for at least a month. I'll be wearing my computer-buying hat in the US and Japan before then. Best regards Adrian Eugenia Kalnay wrote: Dear Adrian: This is a very preliminary response, since my coauthor is away, and I would like to consult with him, and prepare a detailed response/comments, which we will send hopefully next week. I only gave your draft a very fast first reading. Nevertheless, I'll give you a first quick response: First I completely agree with you that ERA-40 should indeed be better than NNR! It better be!!!! So I don't have any quarrel with the results that you present. Hopefully the next global reanalysis will be even better... However, I think that your conclusions about the support or lack of support of your results to those of Kalnay and Cai (2003) are not appropriate or fair: Our method is *totally* based on the fact that the NNR surface temperature was completely independent of the observed surface temperature, which is evidently not true in the ERA-40, where the observed surface air temperatures did influence the soil temperature and moisture. Therefore, the only thing that you can conclude is that our method (unfortunately) cannot be applied "as is" to the ERA-40 reanalysis. For example, the fact that the NNR has a larger amplitude intra-annual variability is in agreement with the assumption that the NNR surface temperatures don't "know" about the increased heat capacity brought by land surface changes. Note also that we do not apply our method to the Rockies, since evidently the NNR surface temperature estimation is not good at all over regions with steep orography and cannot be used. Since you list the criticisms of Trenberth and Vose et al, it would be fair to quote as well our strong responses. (We first had to correct many serious errors of understanding that they made in their first critical comments). We have also received hundreds of very positive comments as well, and one very famous scientist here and his team did a very careful study over China with our method, with very similar results to ours, and with remarkable agreement between the land surface changes estimations from our method, and changes in population and agriculture. There are a number of other points that I would like to make, and Ming Cai will have many insightful comments as well... Thanks for showing this to us first... Please don't send it for publication until we clear out our disagreements, or at least agree to disagree... With warm regards, Eugenia Adrian Simmons wrote: Eugenia I'd appreciate any comments you might have on the attached draft, particularly the second paragraph of the introduction and the last paragraph of the conclusions. I take all the blame - the first the co-authors will see of these two paragraphs are when they respond to the mail below. I hope I have the story straight - I'm certainly open to changing it if not. I also hope that the paper in general does not come across as trying to "do down" the NCEP/NCAR reanalysis - ERA-40 is a much more recent reanalysis, and it would be strange if it did not improve on the NNR in some ways. Here again, I'm open to suggestions for changes in wording. Best regards Adrian -------- Original Message -------- Subject: Paper on ERA-40/CRU comparisons Date: Fri, 23 Apr 2004 17:53:53 +0100 From: Adrian Simmons Reply-To: Adrian.Simmons@ecmwf.int Organization: ECMWF To: Phil Jones , ers@ecmwf.int, erp@ecmwf.int, Anton Beljaars , Vanda Da Costa Bechtold , Sami Saarinen , Pedro Viterbo , Nils Wedi CC: dib@ecmwf.int Phil and ECMWF colleagues Attached is a reasonably polished version of a paper comparing CRU, ECMWF and NCEP analyses of surface air temperature. Many of you have seen a much earlier version. It's in a format that we can put out quickly as an ECMWF project report, after which I think it's worth turning it into a Journal article, which will entail at least dropping a lot of colour from the plots, and probably a few of the time series. I'd appreciate your comments on the paper as it stands for an ERA-40 report, and any suggestions as to what might be cut (or added) to make a paper for ? Journal of Climate. I've tried to address comments received from Phil on an early draft. One specific question for Phil: Do you have the full author list for the Turner et al. paper? I'll send a copy to Eugenia, as I would not want to put this on the web without giving her a chance to respond to the remarks about her letter to Nature with Cai. Are you all happy with the author list? I hope this is not too much of a surprise for Sami and Nils. I think Sami deserves to be there for writing the OI T2m analysis in the first place and for his observational database work for ERA-40, and Nils deserves to be there for getting the AMIP-style simulation done, which I think adds quite a bit to the paper. Best regards Adrian -- -------------------------------------------------- Adrian Simmons Head of Data Division European Centre for Medium-Range Weather Forecasts Shinfield Park, Reading, RG2 9AX, UK Phone: +44 118 949 9700 Fax: +44 118 986 9450 -------------------------------------------------- -- Eugenia Kalnay Distinguished University Professor Department of Meteorology University of Maryland 3431 CSS Building College Park, MD, 20742-2425 tel: 301-405-5370/5391; fax:301-314-9482 ekalnay@atmos.umd.edu; [1]http://www.atmos.umd.edu/~ekalnay Home: Malise Dick and Eugenia Kalnay 56 Lakeside Dr Greenbelt MD 20770 tel/fax 301-313-0208 -- -------------------------------------------------- Adrian Simmons Head of Data Division European Centre for Medium-Range Weather Forecasts Shinfield Park, Reading, RG2 9AX, UK Phone: +44 118 949 9700 Fax: +44 118 986 9450 -------------------------------------------------- Ming Cai Associate Professor Department of Meteorology & School of Computational Science and Information Technology Florida State University Tallahassee, FL 32036 Email: cai@met.fsu.edu, cai@csit.fsu.edu Phone: (850)-645-1551, FAX: (850)-644-9642 -- -------------------------------------------------- Adrian Simmons Head of Data Division European Centre for Medium-Range Weather Forecasts Shinfield Park, Reading, RG2 9AX, UK Phone: +44 118 949 9700 Fax: +44 118 986 9450 -------------------------------------------------- -- -------------------------------------------------- Adrian Simmons Head of Data Division European Centre for Medium-Range Weather Forecasts Shinfield Park, Reading, RG2 9AX, UK Phone: +44 118 949 9700 Fax: +44 118 986 9450 -------------------------------------------------- Prof. Phil Jones Climatic Research Unit Telephone +44 (0) 1603 592090 School of Environmental Sciences Fax +44 (0) 1603 507784 University of East Anglia Norwich Email p.jones@uea.ac.uk NR4 7TJ UK ----------------------------------------------------------------------------