DECISION SUMMARY: coordinate systems used in GLIMS
Bruce Raup
braup at nsidc.org
Thu Apr 4 16:32:45 MST 2002
GLIMSters:
All feedback I received (from 7 or 8 people) regarding storing only lon/lat
coordinates in the GLIMS database was generally positive. So, while we've
been proceeding in this direction already for several weeks now, I would
like to state clearly on this list that this has been firmly decided.
Notes:
1. RCs will do (or are doing) their analyses in the projection of their
choosing. Only final results to be submitted to NSIDC need be converted to
longitude/latitude on the WGS84 datum.
2. It can be (and, indeed, has been) shown that the difference between
interpolation between polygon points in a projection versus along a great
circle is negligible for GLIMS applications.
3. Both local (relative, or within-image) and global (absolute geodetic)
uncertainties will be stored for each polygon or segment. (The new data
dictionary and Entity Relation Diagram will be on the website soon.)
4. GLIMS segments (parts of glacier outlines, centerlines, ELA lines,
etc.) are linked to their source images in the database. Thus, if the
geolocation of a source image is improved upon, the positions of segments
derived from this image (and their uncertainties) can be updated.
5. Having made this decision, the parts of the database design that dealt
with reference points, coordinate system definitions, etc. have been
removed.
6. As has been noted before, there is good and freely available software
for doing coordinate transformations. Examples include:
Proj.4 - Cartographic Projections Library
(http://www.remotesensing.org/proj/). This includes utilities for doing
batch-oriented or interactive coordinate transformations, including datum
shifts.
Generic Mapping Tools (http://gmt.soest.hawaii.edu/)
Cheers to all,
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
More information about the GLIMS
mailing list