date: Fri Nov 13 08:32:28 2009
from: Phil Jones <p.jones@uea.ac.uk>
subject: Re: Land/Ocean
to: Tom Wigley <wigley@ucar.edu>

    Tom,
       I put this in the revised CC essay piece, so have reattached this. If you get a chance
   to look through this, we can probably finish this off in a few iterations.
      Are you getting all my emails?
       GISS use HadISST1 and OIv2.
    OIv2 is this
    Reynolds, R.W., N.A. Rayner, T.M. Smith, D.C. Stokes, and W. Wang, 2002: An
   improved in situ and satellite SST analysis. J. Climate, 15, 1609-1625.
    HadISST1 is a Rayner et al paper - see attached. GISS are essentially using the same SSTs
   as Reanalysis input (what NCEP, ERA-40, ERA-INTERIM and Gil Compo's). So this is different
   from ERSST3b, but I don't think they will differ by much. From 1982 OIv2 must use the same
   SST data as ERSST3b. ERSST3b ought to differ a bit from HadISST1 for years before 1981.  So
   GISS is using spatially infilled SSTs. How they combine these with their 80 boxes from the
   land is unclear. I've looked at the papers and the GISS web site, bt can't see it. They
   have their programs there, so this would be the ultimate way of finding out.
    I'd expect it is something like  Global = a *GISSland + b*SST  where a and b are roughly
   0.3 and 0.7 for the globe.
     Watch for my next email!

    Cheers
    Phil
   At 21:59 12/11/2009, you wrote:

     Phil,
     Can you remind me what GISS uses for SSTs.
     I thought they used NCDC.
     Tom.
     +++++++++
     Phil Jones wrote:

      Tom,
         Sounds fine!
       Still no sight of the HC paper on the 1945-60 adjustments, so nothing also on the
     possible recent changes in SSTs.
        All just confirms what is said in the CC paper that SSTs are what's really important
     for global T.
        On Jim Hurrell's paper about the development of a new SST and Sea-Ice Edge dataset,
     it's not clear whether this has been used by any Reanalysis yet. It hasn't by
     ERA-INTERIM. What was used here is in that paper I sent you a few weeks ago by Adrian
     Simmons. Also clear that you need to put increases in radiative forcing from greenhouse
     gases into these reanalyses. I'll contact Gil Compo again to find out what they used.
         I realized over the weekend that what GISS use for SST in their analysis is the same
     as what was used to force ERA-40 and NCEP. ERA-INTERIM changed this slightly but only in
     the last few years.
      Cheers
      Phil
     At 16:17 08/11/2009, Tom Wigley wrote:

     Content-Type: text/plain; charset=windows-1252; format=flowed
     X-MIME-Autoconverted: from 8bit to quoted-printable by ueamailgate02.uea.ac.uk id
     nA8GI6M5031238
     Phil,
     I have sorted some of this out -- see attached. In fact, this
     is very interesting and kinda cool. The ocean cooling is exactly
     0.2C more than the land cooling ('exact' is partly a fluke of
     course). So this is a nice check on the ocean error. Has this been
     pointed out before?
     Also, the fact that land warming has only begun to accelerate
     relative to the ocean over the past 30 years fits beautifully
     with the suggestion Sarah and I made in 1987 that the early
     20th century warming was a YHC effect. And this all fits with
     the minimal effect of the Sun for this period, as I have noted
     before. I do not this this is generally appreciated.
     I mentioned Hurrell. I'll attach his paper. It is strange
     (sloppy?) that he does not compare the NCAR data set with HadCRU
     or HADSST2. We do need to mention this paper in our Clim Ch
     ms. In fact, we should compare his data with our data.
     Tom.
     Here is my text on the early 20th century warming ...
     "This small solar contribution applies equally to the early 20th century (191040)
     warming. The observed warming over this interval is about 0.5oC (see Figure 1 and Table
     1), while the solar-induced change is either close to zero, perhaps even a cooling
     (assuming no secular TSI component), or a warming of about 0.02oC with a secular
     component (Figure 8). At most, therefore, the solar contribution to early 20th century
     warming is about 4% (even when one assumes a high value for the climate sensitivity).
     This minimal solar effect has been noted previously by Foukal et al. (2004). So, what
     caused the early 20th century warming? A possible explanation is that it is the result
     of a major increase in the rate of formation of NADW (Wigley and Raper, 1987), an idea
     that is supported by the pattern of warming which is a maximum in the North Atlantic
     (Schlesinger and Ramankutty, 1995). We noted earlier that the fact that this warming is
     similar for the land and the ocean (in fact, the 1910 to 1940 trend over the ocean is
     greater than over the land) suggests that it is not externally forced (since this would
     normally lead to warming over the land that was greater than over the ocean), and that
     it originates in the ocean. This also helps to explain why the land/ocean warming
     differential that one would expect as a consequence of external forcing has only become
     evident over the past three decades."

     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
     ----------------------------------------------------------------------------

   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
   ----------------------------------------------------------------------------
