From	Phil Jones	Date	Thu May 5 08:45:53 2005
To	Aiguo Dai, Kevin Trenberth
CC	Jim Renwick, Panmao Zhai, Matilde Rusticucci, David Easterling
Subject	Re: more on section 3.7 and Marengo
Kevin et al, The diagram looks too good to me. CRU's data are reasonable over Brazil for some of the period, but poor in others, particularly recently. So we would have difficulty in updating this because of station numbers and quality. We could try using the GPCC dataset. They have huge numbers of stations for Brazil, but only for specific regions and periods, so likely problems there also. We have a couple of papers in submission to J. Hydrology on flows in the subcatchments of the Parana river, which are well reproduced by rainfall, evaporation and a catchment model. Agree with your concerns about the Amazon flows not agreeing with the rainfall. Do the NAR and SAR regions fully encompass the enormous catchment though. Cheers Phil At 17:36 03/05/2005, Aiguo Dai wrote: One can use the Chen et al. and CRU to produce similar type of plots to validate Marengo's result. He did use the CRU rainfall data set, but not for this particular plot. Aiguo Kevin Trenberth wrote: Hi all As you know we got some manuscripts from Jose Marengo to be considered in our chapter, and he is a LA on another chapter and will be in Beijing. He has offerred to be CA. My question concerns how good his data are? I asked Aiguo Dai to comment: ==== One of the interesting results from Marengo's work is that he found the Northern and Southern Amazonia have opposite phase of decadal rainfall variations (see attached Fig. from Marengo 2004, Ther. Appl. Climatol.): In the northern Amazonia, rainfall is above normal during ~1945-1975 and below normal during ~1976-1998; and it is opposite in the southern Amazonia. He suggested warmer SST in central and eastern Pacific contributed to the dry conditions in the northern Amazonia during 1976-1998. As noted in Betts et al. (2005, JHM, in press), Marengo's basin integrated rainfall index does not correlate well with Amazon river flow during the recent decades (worse than Chen et al.). This large multidecadal signal seems, however, robust. ===== Certainly the attached figure is striking. Are we sure it is not due to changes in the way observations are made? Do other datasets replicate this? The lack of relation with river flow is a substantial concern. Matilde, can you provide informed commentary? If the figure is good then maybe we should include it? Kevin -- Aiguo Dai, Scientist Email: adai@ucar.edu Climate & Global Dynamics Division Phone: 303-497-1357 National Center for Atmospheric Research Fax : 303-497-1333 P.O. Box 3000, Boulder, CO 80307, USA [1]www.cgd.ucar.edu/cas/adai/ Street Address: 1850 Table Mesa Drive, Boulder, CO 80305, USA 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 ---------------------------------------------------------------------------- References 1. http://www.cgd.ucar.edu/cas/adai/