GLIMS feedback as requested

[Luke Copland]

Hi all,
	A few of my thoughts/comments on the email exchange below - Graham's
definition of a glacier makes lots of sense, and will certainly make
their identification easier in areas such as the Canadian Arctic. I
believe that this is an improvement over the original definition as I'm
not a big fan of 'arbitrarily' cutting off tributaries and calling them
separate glaciers as the division is qualitative and open to differences
in interpretation. Bruce's comments that a glacier should only be in one
hydrological basin is a good way of defining glacier boundaries,
particularly on the large scale such as ice caps.
	As Bruce points out, we may ultimately end up with differences in
glacier definition between regions, which really isn't a problem as long
as a future end user can go back and figure out exactly what we've done.
This is where the 'snapshot' images of each glacier in the database can
come in very useful - having an actual satellite image of each glacier
is a whole lot easier to understand (at first) than having a table of
lat/long coordinates. This is particularly true for the public, who may
just be interested in seeing the glaciers which are present in an area
where they are visiting, rather than any of the detailed coordinates.
	
Cheers,
Luke.

-- 
Luke Copland			     Office: Tory 3-20
Dept. Earth & Atmospheric Sciences	Tel: 780 707 5583 
University of Alberta			Fax: 780 492 7598
Edmonton, Alberta T6G 2E3, Canada     Email: luke.copland at ualberta.ca

http://arctic.eas.ualberta.ca/luke

Bruce Raup wrote:
> 
> Graham,
> 
> Thanks for the feedback.  I'll just make a couple of comments.
> 
> On 2001-06-10 00:04 -0400,  J. Graham Cogley wrote:
> 
> >      1. The question asked by Luke Copland still strikes me as pretty
> > fundamental - i.e., what is a glacier? Your answer, that the database
> > should be able to cope with variant definitions, is legitimate but not
> > complete. Variant definitions will tend to limit the synergy that ought
> > to result from a project like GLIMS.  I don't know what the complete
> > answer is, but it might be to adopt the World Glacier Inventory's
> > typology.  Whenever I need a working definition for my own purposes,
> > I use something like "a contiguous collection of complete flowlines
> > through solid ice". This embraces sectors of ice caps as well as valley
> > glaciers, and handles Luke's particular case serenely (the tributaries
> > count as part of the main glacier, but not if they don't make a
> > connection).  The only time I have known it to fail is with Store
> > Supphellebreen in Norway, which consists of a top part feeding a
> > disjunct bottom part by avalanching. (Solution: treat as two glaciers,
> > or stretch the definition of "flowline" a bit.)
> 
> Your definition makes sense.  I take it that this definition prohibits a
> "glacier" from crossing flow divides.  This is what I would prefer.
> 
> We've previously thought that tributary glaciers would be cut off at some
> arbitrary point by the analyst and called a separate glacier.  The reason
> was that many glacier systems have different names for different
> tributaries and are historically viewed as separate glaciers.  However, I
> have no problem with allowing tributaries to be considered part of a
> larger glacier and treated in the database as such, as long as they're in
> the same hydrological basin (which by definition they would be).  The main
> thing we need is temporal consistency on any given glacier.  Consistency
> from region to region might be less important, although it matters quite a
> lot if someone wants to do any study that involves counting glaciers.
> David Bahr's analysis of the fractal nature of the size-number
> distribution of glaciers comes to mind (Water Resources Research, v. 33,
> no. 7, 1997.  "Global Distributions of Glacier Properties:  A Stochastic
> Scaling Paradigm")
> 
> >      NB the definition has to be considered apart from the accuracy
> > of its particular instances.
> >
> >      2. If there is interest in using Luke's glacier on western Axel
> > Heiberg Island (the one with the ASTER DEM) as a test glacier, note
> > that it has already been inventoried once (by Simon Ommanney, using
> > 1959 photos). It would not be a big deal to extract the inventory
> > information pertinent to this ASTER scene.
> >
> >      Best wishes,
> >
> >      Graham.
> >      gcogley@
> >           _/                                            _/
> >          _/                                            _/
> >       _/_/_/     _/ _/_/     _/_/_/      _/ _/_/    _/_/_/    _/     _/
> >        _/       _/_/   _/  _/     _/    _/_/   _/    _/      _/     _/
> >       _/       _/         _/_/_/_/_/   _/      _/   _/      _/     _/
> >      _/  _/   _/          _/          _/      _/   _/  _/  _/     _/
> >      _/_/    _/            _/_/_/    _/      _/    _/_/     _/_/_/  .ca
> >
> >              J. Graham Cogley, Ph.D., Professor of Geography,
> >      Department of Geography, Trent University, Peterborough, Ontario,
> >            CANADA K9J 7B8.  Fax 705-742-2131.  Tel 705-748-1440.
> >                             gcogley at trentu.ca
> 
> Bruce
> 
> --
> Bruce Raup
> National Snow and Ice Data Center                     Phone:  303-492-8814
> University of Colorado, 449 UCB                       Fax:    303-492-2468
> Boulder, CO  80309-0449                            Bruce.Raup at colorado.edu