date: Tue Nov 10 12:40:26 2009 from: Phil Jones subject: Re: Twentieth Century Reanalysis preliminary version 2 data - One to: Gil Compo Gil, One other good plot to do is this. Plot land minus ocean. as a time series. This should stay relatively close until the 1970s. Then the land should start moving away from the ocean. This departure is part of AGW. The rest is in your Co2 increases. Cheers Phil Gil, These will do for my purpose. I won't pass them on. I am looking forward to the draft paper. As you're fully aware you're going to have to go some ways to figuring out what's causing the differences. You will have to go down the sub-sampling, but I don't think it is going to make much difference. The agreement between CRU and GISS is amazing good, as already know. You ought to include the NCDC dataset as well. [1]http://www.ncdc.noaa.gov/oa/climate/research/anomalies/index.html the ERSST3b dataset. In the lower two plots there appear to be two types of differences, clearer in the NH20-70 land domain. The first is when reanl20v2 differs for a single year (like a year in the last 1960s, 1967 or 1968) and then when it differs for about 10 years or so. It is good that it keeps coming back. For individual years there are a couple of years in the first decade of the 20th century (the 1900s). The longer periods are those you've noticed - the 1920s and the 1890s. There is also something up with the period 1955-65 and the 1970s. The 1920s seems to get back then go off again from about 1935 to early 1940s. Best thing to try and isolate some of the reasons would be maps for decades or individual years. For the 1920s I'd expect the differences to be coming from Siberia as opposed to Canada. I think the 1890s might be just down to sparser coverage. The 1890s is the only period where the difference brings your pink line back towards the long-term zero. All the others have the pink line more extreme than the HadCRUT3/GISS average. Rob Allan just called. I briefly mentioned this to him. He suggested maps of data input during these times. He also suggested looking at the spread of the ensembles. Your grey spread is sort of this, but this is a different sort of ensemble to what Rob implied you might have? One final thing - don't worry too much about the 1940-60 period, as I think we'll be changing the SSTs there for 1945-60 and with more digitized data for 1940-45. There is also a tendency for the last 10 years (1996-2005) to drift slightly low - all 3 lines. This may be down to SST issues. Once again thanks for these! Hoping you'll send me a Christmas Present of the draft! Cheers Phil At 20:45 09/11/2009, you wrote: Phil, 1. I didn't get the attached. Both version1 and version2 use HadISST1.1 for SST and sea ice. 2. time-varying CO2, volcanic aerosols, and solar variability (11-year cycle until 1949, "observed" after that) are specified. Attached is a research figure. Please do not share. In it, I have plotted the annual average (top panel) 50S to 70N global average 2m temperature from 20CRv2, SST/2m temperature from HadCRU3, SST/2m temperature from GISTEMP 1200km, and the 90% range of 2m air temperature from 25 CMIP3 models that can be extended beyond their 20C3M runs with SRESA1B. The ensemble mean is the thick gray curve. Averages are July-June. (middle panel) 50S to 70N land-only 2m temperature from 20CRv2, 2m temperature from CRUTEM3, 2m temperature from GISTEMP land-only 1200km. CMIP3 data is the same. (bottom panel) same as middle panel but for Northern Hemisphere land-only (20N to 70N). Anomalies are with respect to 1901-2000. period is July 1891 to June 2005. The CRU (HadCRU) curves are supposed to be black. No data has been masked by another dataset's observational availability, but missing values are not included in that dataset's area-weighted average. Your ERA-Interim finding about it being warmer seems to be the case in the late 19th century but not the early 1920's. Note that the only thermometer data in the magenta curve (20CRv2) is the HadISST1.1 over oceans. The two landonly panels are independent of thermometers, aside from the specified SSTs. There are some very interesting differences, particulary late-19th century, 1920s, and WWII. Correlations (I told you this was research, right?). The second pair is for linearly detrended data. GLOBE (70N-50S) reanl20v2.70n50s.landocean.juljun hadcru3.70n50s.landocean.juljun 0.94370 reanl20v2.70n50s.landocean.juljun hadcru3.70n50s.landocean.juljun 0.82017 reanl20v2.70n50s.landocean.juljun gistemp_combined1200.70n50s.landocean.juljun 0.95284 reanl20v2.70n50s.landocean.juljun gistemp_combined1200.70n50s.landocean.juljun 0.85808 hadcru3.70n50s.landocean.juljun gistemp_combined1200.70n50s.landocean.juljun 0.99088 hadcru3.70n50s.landocean.juljun gistemp_combined1200.70n50s.landocean.juljun 0.97383 GLOBAL LAND (70N-50S) reanl20v2.70n50s.landonly.juljun cru3.70n50s.landonly.juljun 0.85167 reanl20v2.70n50s.landonly.juljun cru3.70n50s.landonly.juljun 0.68755 reanl20v2.70n50s.landonly.juljun gistemp_land1200.70n50s.landonly.juljun 0.81469 reanl20v2.70n50s.landonly.juljun gistemp_land1200.70n50s.landonly.juljun 0.60152 cru3.70n50s.landonly.juljun gistemp_land1200.70n50s.landonly.juljun 0.98050 cru3.70n50s.landonly.juljun gistemp_land1200.70n50s.landonly.juljun 0.95316 NH Land (20N-70N) reanl20v2.nh_nohigh.landonly.juljun cru3.nh_nohigh.landonly.juljun 0.82956 reanl20v2.nh_nohigh.landonly.juljun cru3.nh_nohigh.landonly.juljun 0.67989 reanl20v2.nh_nohigh.landonly.juljun gistemp_land1200.nh_nohigh.landonly.juljun 0.79247 reanl20v2.nh_nohigh.landonly.juljun gistemp_land1200.nh_nohigh.landonly.juljun 0.59900 cru3.nh_nohigh.landonly.juljun gistemp_land1200.nh_nohigh.landonly.juljun 0.98001 cru3.nh_nohigh.landonly.juljun gistemp_land1200.nh_nohigh.landonly.juljun 0.95880 I thought that correlations of 0.8 to 0.85 were high for an independent dataset this long. I think that these are higher than the proxies? The global isn't that fair because we have the HadISST. The correlations are about the same as for AMIP runs, though. See Hoerling M., A. Kumar, J. Eischeid, B. Jha (2008), What is causing the variability in global mean land temperature?, Geophys. Res. Lett., 35, L23712, doi:10.1029/2008GL035984. It will be interesting to see if the masked numbers change. Let me know if you need anything else on this for your essay material. best wishes, gil Phil Jones wrote on 11/9/09 2:55 AM: Gil, A couple of questions. 1. See the attached. Is this paper providing the SST input to 20CRv2? 2. Do you change greenhouse gases in the run? Apologies if these are answered elsewhere. Do you have any pre-draft plots without subsampling to get some idea of how good the agreement? I'm asking these questions as I'm writing an essay for Climate Change. There are no diagrams in this, but showing the agreement with 20CRv2 will be a nice way to finish the paper. Paper briefly documents the magnitude of all the problems in global temperature data - such as SST biases, exposure issues, urbanization and site changes (in order of importance). Site changes for global averages are the least important. Trying to point to a few home truths to skeptics who keep on going on about the land data. Cheers Phil At 15:39 03/11/2009, Gil Compo wrote: Phil, Already calculated. We don't suffer from some of the issues that you and Adrian raised because we use only surface pressure. In the Northern Hemisphere extratropics, the agreement with the various (yours, GISTEMP, NOAA) thermometer-based near surface T is high, but in the Tropics and Southern Hemisphere, there are discrepancies, particularly over Africa and South America. The 20CRv2 does not have the intensity of the Siberia warming. There are also discrepancies in the WWII period. I have not subset the reanalysis to correspond to a particular dataset's missing mask as all 3 have different coverages. I'll be making plots for the paper (with a draft coming) soon. best wishes, gil [2]P.Jones@uea.ac.uk wrote on 11/3/09 3:37 AM: Gil, I'm sitting in a meeting in Bristol with Rob Allan. We've had a thought. When you finish v2 will you be quickly calculating the global T average for the 1891-2006 period? Do you expect this to look like the real global T, or do you expect it to not show the longer timescale change that NCEP from 1948 showed? I can send a paper with Adrian Simmons from JGR in 2004 on this when I'm back in Norwich tomorrow. Cheers Phil Dear Colleagues, Courtesy of the NOAA Earth System Research Laboratory Physical Sciences Division and University of Colorado CIRES Climate Diagnostics Center, at [3]ftp://ftp.cdc.noaa.gov/Datasets/20thC_Rean/provisionalV2/ , please find temporary netCDF files from the 20th Century Reanalysis version 2 (1891-2006). These yearly files are for the ensemble mean analysis (means) and ensemble standard deviation (spreads) of selected variables. Colleagues from organizations contributing to the 20th Century Reanalysis version 2 or the International Surface Pressure Databank version2.2, the observational input dataset, are welcome to investigate these preliminary files. Colleagues on the Atmospheric Circulation Reconstructions over the Earth Working Group 3 Verification and Validation of reanalyses are also welcome to begin working with these files. We are working with our distribution partners at the National Center for Atmospheric Research, the National Oceanic and Atmospheric Administration (NOAA) Earth System Research Laboratory and NOAAs National Climatic Data Center on wider availability and documentation. A rough draft of important documentation is attached. Also, please see our new homepage at [4]http://www.esrl.noaa.gov/psd/data/20thC_Rean/ which includes access to images of 6-hourly sea level pressure and 500 geopotential maps generated from the version 2 data. When production is complete, the 20CR version 2 will span 1871 to present. The references for the dataset are Compo, G.P., J.S. Whitaker, P.D. Sardeshmukh, N. Matsui, R.J. Allan, X. Yin,B.E. Gleason, R.S. Vose, G. Rutledge, P. Bessemoulin, S. Brönnimann, M. Brunet, R.I. Crouthamel, A.N. Grant, P.Y. Groisman, P.D. Jones, M. Kruk, A.C. Kruger, G.J. Marshall, M. Maugeri, H.Y. Mok, Ø. Nordli, T.F. Ross, R.M. Trigo, X.L. Wang, S.D. Woodruff, S.J. Worley, 2009: The Twentieth Century Reanalysis Project. Quarterly J. Roy. Met. Soc., in preparation. Compo, G.P., J.S. Whitaker, P.D. Sardeshmukh, 2008: The 20th Century Reanalysis Project. Third WCRP International Conference on Reanalysis, 28 January 2008, Tokyo, Japan < [5]http://wcrp.ipsl.jussieu.fr/Workshops/Reanalysis2008/Documents/V5-511_ea.pdf >. Compo,G.P., J.S. Whitaker, and P.D. Sardeshmukh, 2006: Feasibility of a 100 year reanalysis using only surface pressure data. Bull. Amer. Met. Soc., 87, 175-190. Whitaker, J.S., G.P.Compo, X. Wei, and T.M. Hamill 2004: Reanalysis without radiosondes using ensemble data assimilation. Mon. Wea. Rev., 132, 1190-1200. Please let us know of any questions about the dataset. And, thank you for your contributions to its development. Best wishes, Gil Compo [6] Jeffrey S. Whitaker [7] 20th Century Reanalysis Project leads -- +++++++++++++++++++++++++++++++++++++++++++++++++++ Gil Compo, Research Scientist, CIRES University of Colorado Mail : CIRES Climate Diagnostics Center NOAA Physical Sciences Division Earth System Research Laboratory 325 Broadway R/PSD1, Boulder, CO 80305-3328 Email: [8]compo@colorado.edu Phone: (303) 497-6115 Fax: (303) 497-6449 [9]http://www.esrl.noaa.gov/psd/people/gilbert.p.compo +++++++++++++++++++++++++++++++++++++++++++++++++++ "Stop and consider the wondrous works of God." Job 37:34 -- +++++++++++++++++++++++++++++++++++++++++++++++++++ Gil Compo, Research Scientist, CIRES University of Colorado Mail : CIRES Climate Diagnostics Center NOAA Physical Sciences Division Earth System Research Laboratory 325 Broadway R/PSD1, Boulder, CO 80305-3328 Email: [10]compo@colorado.edu Phone: (303) 497-6115 Fax: (303) 497-6449 [11]http://www.esrl.noaa.gov/psd/people/gilbert.p.compo +++++++++++++++++++++++++++++++++++++++++++++++++++ "Stop and consider the wondrous works of God." Job 37:34 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 [12]p.jones@uea.ac.uk NR4 7TJ UK ---------------------------------------------------------------------------- -- +++++++++++++++++++++++++++++++++++++++++++++++++++ Gil Compo, Research Scientist, CIRES University of Colorado Mail : CIRES Climate Diagnostics Center NOAA Physical Sciences Division Earth System Research Laboratory 325 Broadway R/PSD1, Boulder, CO 80305-3328 Email: [13]compo@colorado.edu Phone: (303) 497-6115 Fax: (303) 497-6449 [14]http://www.esrl.noaa.gov/psd/people/gilbert.p.compo +++++++++++++++++++++++++++++++++++++++++++++++++++ "Stop and consider the wondrous works of God." Job 37:34 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 ----------------------------------------------------------------------------