ARC - PROJECT - utm -> geographic bug?

[Philip Hurvitz]

One response, so far, from Ken Rukstales <rukstale at gldage.cr.usgs.gov>

KR> Do your coverages cross over 180 degrees longitude?  If so, than 
KR> that is part, if not all, of your problem. 

The answer lies, as usual, in the on-line documentation (as pointed out
by supprt at esri.com):

UTM Limitations
Designed for a scale error not exceeding 0.1 percent within each zone. 
This projection spans the globe from 840 N to 800 S. Error and distortion
increase for regions that span more than one UTM zone. UTM is not designed
for areas that span more than a few zones. See Transverse Mercator for
other limitations. 

TM Limitations
Data on a spheroid or an ellipsoid can not be projected beyond 90 degrees
from the central meridian. In fact, the extent on a spheroid or ellipsoid
should be limited to 15 to 20 degrees on both sides of the central
meridian. Beyond that range, data projected to Transverse Mercator may not
project back to the same position. Data on a sphere does not have these
limitations. 

Because we erroneously chose zone 18, we were way out of the area where we
should have been.  This is more of a projection-mathematics bug, and not a
software bug per se.

Original posting:

> > I've found what appears to be a bug in project.
> > 
> > OS: Solaris & WinNT
> > Ver: 7.1.1, 7.1.2
> > 
> > I've created coverages for large portions of Alaska.  The initial
> > development was done in Mercator porjection.  For more accurate area
> > measurements, these were projected to
> > 
> > Projection                  UTM
> > Zone                         18
> > Datum                      HPGN
> > Units                    METERS             Spheroid GRS1980
> > 
> > But display is better in projection Mercator or geographic, and of course
> > geographic is most flexible for ArcView.
> > 
> > But, arrgh, when I attempt to project these back out to Mercator or
> > geographic, the vectors become corrupted beyond recognition!
> > 
> > I was thinking perhaps it had to do with the coverage boundaries having
> > negative values:
> > 
> >         COVERAGE BOUNDARY
> > Xmin =             -3809313.750            Xmax =             -2590688.750
> > Ymin =              8126191.500            Ymax =             11104837.000
> > 
> > Any clues?
> > 
> > -P.
> > 
> > ***************************************************************************
> >  Phil Hurvitz | GIS Specialist | College of Forest Resources | 355 Bloedel
> > Box 352100 | University of Washington, Seattle, Washington  98195-2100, USA
> >  tel: 206.685.8179 | FAX: 206.685.3091 | e-mail: phurvitz at u.washington.edu
> >              WWW: http://weber.u.washington.edu/~phurvitz/
> > ***************************************************************************
> 
> 
> 
>