date: Fri May 6 16:22:15 2005 from: Phil Jones subject: Re: Section 3.7 to: "David Easterling" Dave, Good on the plots. We'll have something to look at. The DTR plot is already penciled in in 3.2. Phil At 16:01 06/05/2005, you wrote: Phil, I have asked Byron to download the GPCC gridded data and if possible produce a trend map for us to see in Beijing. Also, Jay is going to produce a version of GHCN for the 1948-present period for comparison. I agree that we need to stick with our set periods, and if so then the Chen data set is only valid for the 1979-present period. Plus I just spoke with Russ and he is trying to get us a DTR plot for 1979-present. Dave Phil Jones wrote: Dave, We do need to decide what we show in 3.3 and what comparisons are done in map or time series form. There is also the new GPCC analyses that I sent round details of a few weeks ago. Aiguo Dai sent some time series comparisons a week or so ago with these in and highlighted a number of issues. I've had some email discussion with GPCC on this and I'll bring this with me. A lot of the detail you refer to will likely have to go into one of the Appendices. 3.2 may be a useful guide here with time series comparisons of the various datasets, but only one for maps of trends. Temperature is a lot easier, but if we set periods, we may be able to follow those 'rules'. Periods are a problem, but I think we should be sticking to just the 2, from 1901 and 1979. Before adding Figures, we need to decide which we really need as we have too many. We are 50% oversize. Kevin, Jim and others have reduced some of 3.5-3.7 and also done some work on 3.4. David, Kevin and me have worked on 3.2 and the Appendices. But all this hasn't saved that many pages so far (< 5). References will have gone down by a page or two. Figures are where we are likely to make most reductions. I know we won't manage 50%, but 25% is possible. Hope you all have good trips to Beijing. See you all next week ! Phil At 13:43 06/05/2005, David Easterling wrote: All, I too am a bit concerned about results from the various pcp datasets. I think this really warrants a discussion next week in Beijing. Part of the problem I see is that trend results are strongly impacted by the starting and ending points of the time series and we have different trends with different starting and stopping points. The Chen et al dataset is 1948-2003, GHCN has both 1901-2003 and 1979-2003 in our plots, GPCP has 1979-2003 and so on. We are going to produce a GHCN plot for the 1948-2003 period for comparison with Chen. Also, they have very different methodologies in their development. Chen uses optimal interpolation of gauge data and and EOF reconstruction over the ocean but is based on both GHCN and synoptic reports from the NOAA/CPC Climate Anomaly Monitoring System (CAMS) dataset, GHCN is only land-based gauge data with a simple gridding scheme but some of the data (e.g. Canada and Russia) have been adjusted for undercatch, etc., and GPCP is a blend of satellite and gauge data. Also we need to be careful when we use the word "trend" and probably should use something like "change" in many places. Even the global temperature time series since 1900 does not exihibit a monotonic trend, but has three distinct phases, increase, no change, then increase. Dave Easterling Kevin Trenberth wrote: Hi all, and especially Panmao, Jim and Matilde (who I challenged with having a go at revising 3.7). Jim has made a good attack on 3.5 an 3.6 and progress is being made there. However, perhaps the section where the biggest shortcomings exist are 3.7. It is too long, and not focussed enough on the topic of change in monsoons, etc. As I read through it, 3.7.1 is mostly about changes in rainfall, and so this MUST be fully compatible with 3.3 and it should not repeat what is in 3.3. Instead it should refer to 3.3 and some 3.3 figures might be modified to accommodate 3.7 and monsoon needs. Accordingly, I want to recruit David Easterling to also look hard at 3.7. I am quite concerned about the multiplicity oif rainfall datasets and the sometimes different stories they tell, and the lack of clear timelines. Saying something is a trend for 20 years and then it changed sign means it is not a trend. The discussion should take place more in the context of how we think things should or might change. In monsoon areas, there are strong contrasts of rain areas and subsidence, and if the circulation gets stronger then wouldn't both get more intense? Doing area averages over regions that involve both make no sense. Would shifts be more likely: dipole structures? With land warming faster than oceans, does that mean enhanced monsoons? Given pollution (dimming), wouldn't rainfall be reduced in some areas? Or actually what that really does is change the frequency and intensity of rainfall, making rains less frequent, but still heavy because there is more water vapor. Is there evidence of this? Does this have a seasonal signal: mainly in winter but not the monsoon season? Can we answer these kinds of questions? Can we say how monsoons are changing? Are they all changing in the same ways or are they different? Can we bring the common features up front? Analyses of rainfall data or CAPE (as in the Australian section) are not confined to Australia. Note my comment that the DeMott and Randel (2004) paper is likely contaminated by the erroneous corrections applied to some sonde data that have since been corrected (Durre et al 2002). This intersects with 3.4.1 I decided to go through the section with a "blunt ax" approach and cut stuff that seemed not obviously pertinent. Even then I probably did not go far enough. I stress that this has NOT taken into account suggestions from reviewers. But I thought I would do this as a challenge to you to say why this stuff should not be cut? i.e. often the default is to leave it the way it is unless critical comments suggest otherwise. In this case I want to suggest that at least for the cut material, that the default be to leave it out unless justified. It is not that some of the cut material is wrong, it is just that it is focussed on mean conditions or ENSO relations that are dealt with elsewhere in our chapter (3.6), or other material that is not adequately focussed. This section is supposed to deal with the phenomenology of monsoons, Hadley and Walker circulation, etc. It should not repeat what is in 3.3. It also has the chance to regionalize some aspects, which is why I am including all LAs on this email. In case you still have not quite got the message, I wish to provoke you into action on this section. Don't get upset with what I have done, instead get on with it! Bring your suggestions to Beijing. Thanks Kevin -- **************** Kevin E. Trenberth e-mail: [1]trenbert@ucar.edu Climate Analysis Section, NCAR [2]www.cgd.ucar.edu/cas/ P. O. Box 3000, (303) 497 1318 Boulder, CO 80307 (303) 497 1333 (fax) Street address: 1850 Table Mesa Drive, Boulder, CO 80303 -- David R. Easterling, Ph.D. Chief, Scientific Services Division NOAA's National Climatic Data Center 151 Patton Avenue Asheville, NC 28801 USA V: 828-271-4675 F: 828-271-4328 [3]David.Easterling@noaa.gov 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 [4]p.jones@uea.ac.uk NR4 7TJ UK ---------------------------------------------------------------------------- -- David R. Easterling, Ph.D. Chief, Scientific Services Division NOAA's National Climatic Data Center 151 Patton Avenue Asheville, NC 28801 USA V: 828-271-4675 F: 828-271-4328 [5]David.Easterling@noaa.gov 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 ----------------------------------------------------------------------------