date: Mon, 15 Jan 2007 13:55:14 UT
from: grlonline@agu.org
subject: Review Received by Geophysical Research Letters
to: p.jones@uea.ac.uk

   Content-Transfer-Encoding: binary Content-Type: multipart/alternative;
   boundary="_----------=_116886931419765" MIME-Version: 1.0 X-Mailer: MIME::Lite 3.01 (F2.6;
   B2.11; Q2.03) Date: Mon, 15 Jan 2007 13:55:14 UT Message-Id: <116886931486@gems2>
   Dear Dr. Jones:
   Thank you for your review of "Warming in the High Arctic: Evidence from an instrumental
   record spanning 125 years" by G.W.K. Moore and Aarti Motala [Paper #2006GL028883], which we
   have safely received. A copy of this review is attached below for your reference.
   Thank you for your time and effort!
   Sincerely,
   Mark New
   Editor
   Geophysical Research Letters
   ------------------------------------------------------------------------
   Science Category: Science Category 4
   Presentation Category: Presentation Category C
   Annotated Manuscript: No
   Anonymous: Yes
   Referrals: No
   Confidential Referrals:
   Highlight: No
   Highlight:
   Little
   Formal Review:
   Review of Moore and Motala
   General View
   This paper has some potential, but needs a major rewrite in order to be worthwhile. As I
   see this as a very major revision, I'm recommending rejection, but suggesting that the
   author's submit a massively revised version. I have come to this conclusion for several
   reasons, outlined in the major changes section below. Although the authors acknowledge the
   assistance of four colleagues, they have shown themselves (particularly with regard to
   reanalyses) to be relatively nave and not fully aware of the climatological literature.
   1. The instrumental record at Alert extends back to 1950/51. You cannot ignore 29 years of
   modern instrumental data. You are only using monthly mean temperatures and these data are
   readily available. With the Arctic showing recent warming, larger-scale averages indicate
   warming since the 1960s and a warm epoch in the 1930s and 1940s. Either way there is a lot
   more data - of a potentially cooler period, for which modern measurements exist. The
   earlier expedition/exploration data might still be unusual but it is vital to also show the
   full Alert record (the most northern 'manned' site in the world) compared to the full
   Arctic record for the land areas north of 65N.
   2. The aim of the paper was to compare the late 19th century years with the full Alert
   record on Ellesmere Island. My next point will discuss the poorest part of the paper - the
   use of reanalyses - but there is a much simpler way of assessing the unusualness of the
   early data. There are a number of series from Greenland, which extend back to the period
   and some to much earlier periods (see Vinther et al., 2006). Apart from not being aware of
   this work, the arguments used by the authors from Rigor et al. (2000), especially for the
   winter season where the authors' differences are largest, mean that they can make a direct
   comparison with the homogenized record from Upernavik (~73N). I realise this site is
   someway south, but correlation decay lengths are large in winter. The winters around the
   expedition period do appear cool, but there are other winters that are much milder. The NW
   Greenland record shows considerable decadal variability in winter.
   3. To use reanalyses like this there needs to be more awareness of what input data are
   used. If the regional reanalysis of Mesinger et al. (2006) is based on the same system as
   NCEP (NNR, i.e. Kalnay et al., 1996), then the whole exercise is pretty pointless. The
   authors should read the paper by Simmons et al. (2004) where the quality of the ERA-40
   reanalysis is compared with observational data and NNR. ERA-40 is a second generation
   reanalysis and the surface temperature measurements are assimilated. This is not the case
   in NNR. This might explain why the real observed data at Alert only are only poorly
   explained by the regional reanalysis. I think the use of the reanalysis in this instance is
   pointless and I remain to be convinced that their use serves any purpose or that they are
   giving any useful information as to the comparison of temperatures measured at Alert and
   the expedition sites.
   4. When using reanalysis data for comparison to real observations near coastal areas, it is
   vital to know whether the model grid box being used is land or ocean. This may not matter
   much for most of the year, but may in summer months. The resolution of the regional
   reanalysis needs to be stated and perhaps the grid-box size shown on a figure, such as
   Figure 2. What is the resolution?
   5. The discussion of exposure issues on p4 needs to bear in mind where the observations
   were made and when. These papers about exposure need to be read. Stevenson screens and
   their variants were in widespread use in the UK, Canada and the USA in the 1880s. No north
   European location has problems due to exposure in this period. Problems come earlier -
   prior to the 1860s - see discussion in Moberg et al. (2003). The warm bias in summer is
   clear for European locations, but that in winter is much less marked. The biases of the
   order of 0.2C are an order of magnitude smaller than those that might be due to different
   locations on Ellesmere Island. The authors are right to raise the issue, but it is likely
   unimportant, 0.2C is likely the wrong value, but the real one isn't known.
   6. The authors should comment on the best way to make use of the early expedition records
   like those being analyzed. This is to use the approach of Klingbjer and Moberg (2003).
   Here, early measurements some distance from a modern station were adjusted by means of
   modern measurements taken with a data logger. This might have been impossible, but is
   relatively unexpensive, just requiring a little forethought.
   7. Two small final points. The Greenland data will show whether these years in the 1870s
   and 1880s are unusual. Earlier north Canadian expedition data have been assessed by
   Overland and Wood (2003) and found not to be that unusual compared to modern conditions.
   This was for different periods and different regions though. It may just show that you
   can't take as representative a few years of measurements.
   8. Related work to that discussed in the authors has been undertaken in the Antarctic with
   expeditions and exploration data (see Jones, 1990). There are likely other datastes in
   other parts of northern Canada, which along with the much earlier Greenland data could be
   combined in a far more informative paper. The paper is timely with respect to the upcoming
   IPY, but there were many more Arctic expeditions in 1882/3 during the first IPY. Many of
   these have been recently reconsidered in the light of the upcoming period.
   9. There may also be some proxy data for Ellesmere Island from lake cores, but these mostly
   reflect summer conditions. Ray Bradley has written frequently on this area.
   References (not used by the authors)
   Klingbjer, P. and Moberg, A., 2003: A composite monthly temperature record from Tornedalen
   in northern Sweden, 1802-2002, Int. J. Climatol. 23, 1465-1494.
   Moberg, A., Alexandersson, H., Bergstrm, H. and Jones, P.D., 2003: Were south Swedish
   summer temperatures before 1860 as warm as measured?, Int. J. Climatol. 23, 1495-1521.
   Overland, J.E., and K. Wood, 2003: Accounts from 19th century Canadian Arctic explorers'
   logs reflect present climate conditions, EOS, 84, 410-412.
   Simmons, A.J., P.D. Jones, V. da Costa Bechtold, A.C.M. Beljaars, P.W. Kllberg, S.
   Saarinen, S.M. Uppala, P. Viterbo and N. Wedi, 2004: Comparison of trends and low-frequency
   variability in CRU, ERA-40 and NCEP/NCAR analyses of surface air temperature. J. Geophys.
   Res., 109, D24115, doi:10.1029/2004JD006306.
   Vinther, B.M., Andersen, K.K., Jones, P.D., Briffa, K.R. and Cappelen, J., 2006: Extending
   Greenland temperature records into the late-18th century. J. Geophys. Res. 111, D11105,
   doi:10.1029/2005JD006810.
