date: Thu Feb 28 14:46:03 2002
from: Keith Briffa <k.briffa@uea.ac.uk>
subject: Re: review of Klok and Oerlemans
to: r.braithwaite@man.ac.uk

   thannnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnkyou
   Keith
   At 02:39 PM 2/28/02 +0000, you wrote:

     Review of:
     DERIVING A CLIMATE HISTORY FROM A GLACIER LENGTH RECORD BY LINEAR
     INVERSE MODELLING
     Klok and Oerlemans
     Submitted to The Holocene
     The paper says (p. 3) that a "climate signal" can be extracted from a
     glacier length record. However, the "climate signal" is obviously a
     mass balance history or a reconstruction of equilibrium line
     altitude (ELA). This make me wonder if we are really talking about a
     "climate signal" or "climate history". I would expect something about
     temperature and precipitation here.
     The median elevation of a glacier (p. 3) according to Manley (1959)
     is not the real median. The simple methods to reconstruct ELA
     were never intended to consider or include response time. There are
     certainly more methods than listed here, e.g. see Benn and Lehmkuhl
     (2000).
     The author admits that a numerical flowline model (p. 3) would be
     best but says that their aim is to develop a simple "analytical"
     model that should be more applicable. I am not sure what they mean
     with "analytical". Other terminology is a problem. For example, they
     define "climate sensitivity" in terms of the change in glacier length
     with a change in the ELA, while this term is already well understood
     by climate modellers. Equations (1) and (2) look quite scientific but
     they don't really express much more than the fact that almost any
     phenomenon can be described by a Taylor series, truncated after the
     first term.
     I have real problems calculating time derivatives with fitted
     polynomials and filtered data (p. 6).
     On p. 7 and p. 11  they cite an non-existent publication (Oerlemans,
     2001) which appears to provide important background.
     On p. 8, nu (a Greek letter that I can't print in my mail program) is
     defined as a ratio of mean glacier thickness to glacier length but is
     taken as zero on p. 9. How can that be?
     I can't help thinking that the authors make a bit of a mystery about
     their "analytical" model. A simple conceptual relation between
     glacier length change and change of ELA was proposed by Callendar
     (1950). As this relation includes two different glacier widths (at
     the snout and at the ELA) and slope at the ELA it is a reasonably
     sophisticated predecessor of their model. The authors should at least
     reference Callendar's model.
     The approach is interesting and certainly deserves publication in the
     Holocene but I don't find this a very reader-friendly paper. Pages
     12-17 seem very discursive and approach is rather unsystematic. The
     authors should tighten up the text and even consider changes of
     structure.
     REFERENCE
     Benn, D. I. and F. Lehmkuhl. 2000. Mass balance and equilibrium-line
     altitudes of glaciers in high-mountain environments. Quaternary
     International 65/66, 15-29.
     Callendar, G. S. 1950. Note on the relation between the height of
     the firn line and dimensions of a glacier. Journal of Glaciology
     1(8), 459-461.
     *****************************************
     *  Roger J. Braithwaite Ph.D.
     *  Reader in Physical Geography
     *  School of Geography
     *  The University of Manchester
     *  Oxford Road
     *  Manchester M13 9PL
     *  U.K.
     *  Tel: 0044-161 275 3644 (direct line/answerphone)
     *  Tel: 0044-161 275 3636 (Dept. Office/messages)
     *  Tel: 0044-1663 745 873 (home/answerphone)
     *  Fax: 0044-161 275 7878
     *
     [1]http://www.art.man.ac.uk/geog/home.htm

     *****************************************

   --
   Professor Keith Briffa,
   Climatic Research Unit
   University of East Anglia
   Norwich, NR4 7TJ, U.K.

   Phone: +44-1603-593909
   Fax: +44-1603-507784
   [2]http://www.cru.uea.ac.uk/cru/people/briffa[3]/

