Generation of glacier IDs for GLIMS: perl code

[Bruce Raup]

Hi all,

In addition to messages pointing out my north-south mistake, I got some
feedback (much of it from Hugh Kieffer) on GLIMS glacier IDs:

1.  It was thought that having a flexible number of digits to the right of
the implied decimal place could lead to confusion.  Thus, leaving it at 3
digits was recommended.

2.  It was pointed out that the underscore is unneeded.

3.  There was a preference for having the E/W, N/S designators *follow*
their corresponding numbers.

4.  The original specification called for only east longitudes, in the
range of [0,360).  Luke Copland stated that using west longitudes for his
area was easier.  If we allow W or E longitudes to be used, then we must
allow longitudes only in the range (-180,180].  Otherwise, we could get
two IDs in the western hemisphere that point to the same place on the
ground, but that would be distinct IDs (e.g. G050000W040000N and
G310000E040000N).

Reminder:  the rationale for having the lon/lat encoded in the glacier ID
is so that Regional Centers can generate unique IDs without having to
consult with anyone else (like, e.g. querying the database for the next
available ID).

So, I propose an ID format as follows:

  Gdddddd[W|E]ddddd[N|S]

There are 3 digits to the right of the implied decimal place.  [W|E] means
either W or E.  Longitudes must be in the range [0,180], with W or E
designating hemisphere.  Latitudes range [0,90].  As an example, if there
were a glacier where NSIDC is, it would have the ID "G105253W040013N".

IS THIS OKAY WITH EVERYONE?

Thanks,
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