date: Tue Oct 24 16:06:31 2000
from: Phil Jones <p.jones@uea.ac.uk>
subject: Re: Norwegian Norpast project
to: <Oyvind.Nordli@dnmi.no>


 Dear Oyvind,
    First, in response to your questions.

 1. I'm attaching a file of MSLP reconstructions from pressure data for
 the greater European area. You need to look at the paper Jones et al
 (1999) Int. J. Climatol. 19, 347-364. The file gives these reconstructions
 for 1780 to 1995 using the method outlined in the paper. There are 60
 grid points, ordered by grid point. After each monthly MSLP value there
 is number after a /. This number is a quality code depending on how good
 the reconstructions were. The scale is 0-4 , with 0 being the best.
 This assessment is discussed in the paper Jones et al (1987) referred to
 in the above paper.  Let me know if you have problems with the file. I
 hope it transfers OK.
    There are earlier reconstructions (see Luterbacher et al., 2000 in
 IJC 20, 1049-1066). These however are based on data other than MSLP
 (eg temp and precip series) so there is a clear case for circularity
 when looking at circulation/climate relationships as they are used in
 the reconstructions. The 1999 paper is based solely on MSLP data.

 2. I know of no GCM runs that have been used for the whole of the
 Holocene. The best that I know of are runs with estimates of past
 solar variability and volcanic forcing from about 1700. The Hadley Centre
 and MPI have run such models, but I don't have the results. The earlier
 a model run is started the more difficult it is to set the boundary
 conditions - namely the SST's everywhere. It is likely that the runs
 that have been made have started with SST's that are not right, not just
 SSTs but also temperatures at depth in the ocean.

   I have been looking at some long temperature records from Fennoscandia,
 particularly those put together during the IMPROVE project. This involved
 Anders Moberg in Stockholm and Hans Bergstrom in Uppsala. Summer
 temperatures for both these series which extend back to 1756 and about 1740
 show warmer conditions than today during many summers from 1750 to 1850.
 Uppsala has clear problems between 1722-1740 so these data have been
 ignored. Warm summers at this time also occur at St. Petersburg (also in
 IMPROVE), so I partly take the view that the early records from Trondheim,
 Vardo (from 1830s) and Archangel (from 1810s) may be correct. Winters
 in all the series were much colder than today at these times. Continentality
 may have increased during these times (greater winter/summer contrasts).
 Exposure may still be a problem, but Anders believes this isn't an
 important factor. Some central European series in Germany and Austria
 show similar summer/winter trends but not quite to the same extent as
 in Fennoscandia. Central England also shows little change in summer
 since the 18th century. Autumns for CET in the 1730s were clearly warmer
 than today. All somewhat perplexing.
    If you have any very early SST or pier sea temperature measurements
 this may help.

 Best Regards

 Phil



At 02:18 PM 10/24/00 +0200, you wrote:
>Dear Phil Jones,
>
>I work at the Norwegian Norpast project on historical climatology. The
>sources I have used so far are historical proxy data, particularly the start
>of the grain harvest, which correlate very well with mean spring/summer
>temperature. The data go back to the mid 18th century, e.g. to the little
>ice age. I am a little confused concerning the old Scandinavian instrumental
>series that I suspect have false trends in summer due to radiation errors of
>the thermometers.
>
>The proxies procured within Norpast will be used for temperature
>reconstruction in different Norwegian regions. In addition to the already
>mentioned harvest data, there are also tree-rings, pollen, chironomids and
>others. I need to interpret the results, ensuring that the picture I get of
>the climate variations and trends from the little ice age to the present
>climate is consistent. The main mountain ridge that separates the eastern
>and western parts of Norway has a large influence on regional temperature
>even in summer. Thus, spatial temperature patterns are linked to
>circulation.
>
>So my questions are:
>
>1)  Is it possible to get monthly SLP maps for the 18th and 19th centuries
>from UEA, and eventually which period is available. If there exists gridded
>versions of the maps, this would be much preferable.
>
>Within Norpast we also study the climate of the whole Holocene, and we see
>interesting differences in the extent of the glaciers in different parts of
>Norway. They are sometimes not in phase. We plan to look closer at this, and
>think that empirical downscaling/upscaling techniques could be used to link
>historic climate variations to atmospheric circulation. In doing so we need
>to co-operate with a model community that have large scale GCM runs for the
>Holocene.
>
>2) Do you have any suggestions where we could get results from a GCM-run
>covering Holocene that could be used for empirical downscaling in the area
>of the Norwegian sea and Scandinavia?
>
>Kindest regards from
>
>
>Øyvind Nordli
>
>
>
>
>
>
>E-mail: oyvind.nordli@dnmi.no
>Phone: + 47 22 96 32 61
>Post:
>Øyvind Nordli
>Norwegian Meteorological Institute
>Box 43 Blindern
>N-0313 Oslo
>
>
>