GLIMS definition issues

[Bruce Raup]

Hello all.

Thanks to Jeff, Hugh, Roger, John Dwyer, and Richie for your responses.
I hope to get more.

1.  I perhaps should have been clearer what I meant by "definition of a
    glacier".  I meant it in the context of the database only -- not the
    full scientific definition.

2.  Ice shelves:  The consensus so far seems to be that we should include
    ice shelves in the GLIMS database.  I agree.

    However, for ease of processing in the database, I propose that
    outlines of glaciers that flow into ice shelves end at the grounding
    zone [1], and that ice shelves have their own outlines.  This will
    simplify treatment of ice shelves which have many glaciers flowing into
    them (which is the common case), and it avoids having to place the
    glacier outline along flow stripes in the ice shelf, which may or may
    not be visible.

    Ice tongues should probably be treated as part of the parent glacier.
    For example, the ice tongue at the terminus of the Mertz glacier should
    probably be included in the Mertz Glacier outline.  Does that sound
    reasonable?  However, this approach may cause a slight problem when the
    ice tongue has a different name, as Drygalski Ice Tongue / David
    Glacier.

3.  Coordinate systems:  First, a little clarification.  The original
    concept behind the N/E system was to explicitly separate relative
    uncertainty (precision) from absolute uncertainty (accuracy).  In a N/E
    system, all coordinates are relative to something you can identify in
    imagery.  That system is placed in an absolute (geodetic) system by
    knowing the Lon/Lat of the origin of the N/E system.

    A few RCs have expressed the view that N/E systems won't work well for
    them, especially in dealing with large areas where all glaciers must be
    treated in the same coordinate system (for mosaicking and the like).
    They would prefer to use lon/lat directly (or, I suppose, UTM, but both
    are absolute, rather than relative, systems).

    Thus, in order to satisfy the needs of these RCs as well as the need to
    separate precision from accuracy, I tried to state that the database is
    capable of storing information in either N/E or L/L, and that if L/L
    was used, the precision/accuracy separation must be made explicit in
    the database.

    Based on your comments, it is apparent that I need to make my writing
    clearer before submitting the manuscript to Eos.  :-)

[1]  "Grounding zone" is a term I've heard used by researchers attempting
     to locate them (e.g. Neal Young).  I've assumed it was because their
     location was fuzzy, either in time (due to tides?) or space (due to a
     zone of spatially intermittent grounding?).  If one term is better
     than the other, please let me know.

Thanks for your continued input.

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                                    braup at nsidc.org