cc: Nick McKay <nmckay@email.arizona.edu>, Caspar Ammann <ammann@ucar.edu>, Bradley Ray <rbradley@geo.umass.edu>, Keith Briffa <k.briffa@uea.ac.uk>, Miller Giff <gmiller@colorado.edu>, Otto-Bleisner Bette <ottobli@ucar.edu>, Overpeck Jonathan <jto@u.arizona.edu>, Bo Vinther <bo@gfy.ku.dk>
date: Tue, 2 Jun 2009 14:43:50 -0700
from: Darrell Kaufman <Darrell.Kaufman@nau.edu>
subject: Re: a few comments on the latest draft
to: David Schneider <dschneid@ucar.edu>

   Dave:

   Thanks for your helpful suggestions.

   (1) Your first comment refers to a key statement in the abstract and we need to make sure
   it's crystal clear. We need to somehow quantify the extent of recent warming in context of
   the last 2000 years.  Would it be cleaner to just say, "based on our reconstruction, the
   1990s was the warmest decade of the first 200 decades AD?". Or, "the last half-century
   (1950-2000 AD) included four of the five warmest decades of the last two millennia".

   These are easier to grasp, but they do not include two important aspects:

   - our main conclusion that the overall cooling trend makes the recent warming
   even more unusual than previously assumed (granted, not by much).

   -  the exceptional warming that has taken place in the last 10 years; our reconstruction
   ends at 1999, so significantly under-represents the warming that has taken place.

   The current statement "During the most-recent decade, temperatures rose 1.4°C higher than
   the projected value" takes into account both the projected cooling and the 21st century
   warming.

   I'm open to suggestions. What do others think?

   (2) I have not done any correlations or spectral analysis to compare our Arctic
   reconstruction to other Hemispheric averages. It would be great if someone wants to tackle
   this for either this paper or in the future. I have all of the data in a tidy spreadsheet.

   (3) I have now added: "General Comments: The reviewer was concerned that the relation
   between orbital forcing and temperature response is stronger in the proxy data than in the
   climate simulation. Reviewers comment #10 included additional helpful suggestions for
   improving the data-model comparison.  We have re-analyzed the CCSM model output to include
   a full 2000-year-long transient simulation. This places the proxy data on par with the
   model data, and led to improved regression statistics, which now show more convincingly
   that the climate sensitivity of the CCSM is in good agreement with the sensitivity of the
   proxy reconstruction."

   Thanks,

   Darrell

   On Jun 2, 2009, at 1:51 PM, David Schneider wrote:

     Darrell et al.:

   Just a few comments on the latest draft:

   1) Lines 25-27: "temperatures rising 1.4 degrees higher than projected..." is confusing.
   When I first read this, I thought you are referring to model simulations. Having read the
   rest of the text, it becomes clear you're referring to projecting the linear trend from the
   proxy data forward. Can you make this clearer in the abstract? By the way, the 20thC model
   simulations show warm season temperatures rising about 1.5 degrees from about 1900 to 2000,
   for example: [1]http://www.cgd.ucar.edu/%7Edschneid/areatimeseries1.gif

   2) In Figure 3 F, you show 'this study' compared with Moberg, Mann et al., etc. Are there
   any good correlations, spectral coherence, etc. between this study and any of the others?
   This might bolster the arguments made in lines 135-145.

   3) I think the reply to Reviewer 2 needs a general comment about the revised analysis
   showing the climate sensitivity of the CCSM to be in good agreement with the proxy-inferred
   sensitivity. And better regression stats than before, thanks to his helpful suggestions.

   Dave