[Geotiff] Linear Units of Projections Parameters

[Bunch, Bryan (EPPC OOS OIS)]

I have encountered behavior with ESRI ArcGIS/ArcObjects components that
indicates an automatic assumption of meters for false origin values
regardless of what is specified by ProjLinearUnitsGeoKey.  If I am not
mistaken the ESRI components that interpret and write geokeys are
actually provided by ERDAS, which in turn may be utilizing libteotiff
(I'm really not sure).  Perhaps the raster teams at ESRI or ERDAS
monitor this list and could clarify my observation.

Either way, this was a serious problem for me when I had to define the
Kentucky Single Zone SPCS within our geotiff products since the GeoTiff
Standards is out of date and does not include KY1Z even though the
latest version of the EPSG dataset does (see
http://bugzilla.remotesensing.org/show_bug.cgi?id=989).

Here is how I wound up defining the KY1Z projection in our geotiff
products (e.g. http://kymartian.ky.gov/krgmaps) in order to get it to
work with the various interpretations we have encountered:

Geotiff_Information:
   Version: 1
   Key_Revision: 1.0
   Tagged_Information:
      ModelTiepointTag (2,3):
         0.5              0.5              0
         5567330          3987430          0
      ModelPixelScaleTag (1,3):
         6.66666667       -6.66666667      0
      End_Of_Tags.
   Keyed_Information:
      GTModelTypeGeoKey (Short,1): ModelTypeProjected
      GTRasterTypeGeoKey (Short,1): RasterPixelIsArea
      GTCitationGeoKey (Ascii,81): "snipped"
      GeographicTypeGeoKey (Short,1): GCS_NAD83
      ProjectedCSTypeGeoKey (Short,1): User-Defined
      PCSCitationGeoKey (Ascii,70): "snipped"
      ProjectionGeoKey (Short,1): User-Defined
      ProjCoordTransGeoKey (Short,1): CT_LambertConfConic_2SP
      ProjLinearUnitsGeoKey (Short,1): Linear_Foot_US_Survey
      ProjStdParallel1GeoKey (Double,1): 37.0833333
      ProjStdParallel2GeoKey (Double,1): 38.6666667
      ProjNatOriginLongGeoKey (Double,1): -85.75
      ProjFalseEastingGeoKey (Double,1): 4921250
      ProjFalseNorthingGeoKey (Double,1): 3280833.33
      ProjFalseOriginLongGeoKey (Double,1): -85.75
      ProjFalseOriginLatGeoKey (Double,1): 36.3333333
      ProjFalseOriginEastingGeoKey (Double,1): 1500000
      ProjFalseOriginNorthingGeoKey (Double,1): 1000000
      End_Of_Keys.

   End_Of_Geotiff.

Projection Method: CT_LambertConfConic_2SP
   ProjFalseOriginLatGeoKey: 36.333333 ( 36d20' 0.00"N)
   ProjFalseOriginLongGeoKey: -85.750000 ( 85d45' 0.00"W)
   ProjStdParallel1GeoKey: 37.083333 ( 37d 5' 0.00"N)
   ProjStdParallel2GeoKey: 38.666667 ( 38d40' 0.00"N)
   ProjFalseEastingGeoKey: 4921250.000000 m
   ProjFalseNorthingGeoKey: 3280833.333333 m
GCS: 4269/NAD83
Datum: 6269/North American Datum 1983
Ellipsoid: 7019/GRS 1980 (6378137.00,6356752.31)
Prime Meridian: 8901/Greenwich (0.000000/  0d 0' 0.00"E)
Projection Linear Units: 9003/(unknown) (1.000000m)


Note that the ProjFalseEastingGeoKey and ProjFalseNorthingGeoKey are
defined in US Survey Feet (4921250 and 3280833.33... respectively) while
the ProjFalseOriginEastingGeoKey and ProjFalseOriginNorthingGeoKey are
defined in meters (1500000 and 1000000 respectively).  I had to include
both sets of parameters in the definition because the images were
rectified with transformation parameters (cell size and origin) defined
in US Survey Feet, not meters.  So far, this scheme has worked, but it
would be much better if all software followed the guidelines outlined in
the GeoTiff Specification and honored the units specified by the
ProjLinearUnitsGeoKey.

By the way, note how the ProjFalseEastingGeoKey and
ProjFalseNorthingGeoKey values are presented in the Projection Method
section above.  The values given are in US Survey Feet while they are
labeled as (m).  This is also covered in the bugzilla report at
http://bugzilla.remotesensing.org/show_bug.cgi?id=989.

Bryan


-----Original Message-----
From: geotiff-bounces at lists.maptools.org
[mailto:geotiff-bounces at lists.maptools.org] On Behalf Of Frank Warmerdam
Sent: Saturday, February 25, 2006 10:22 AM
To: geotiff at lists.maptools.org
Subject: [Geotiff] Linear Units of Projections Parameters


Folks,

What are peoples opinions on how the units of linear quantities like
false easting and northing should be interpreted in a GeoTIFF file?

Reviewing the specification (ie. the definition of
ProjFalseEastingGeoKey in 2.7.2 of the specification) it seems clear
that the false easting and northing should be in the units defined by
ProjLinearGeoKey.

But listgeo's normalized parameter report, GDAL, and a variety of other
commercial software packages all seem to make the assumption that false
easting and northing are always in meters.

I have created a bug report against the specification about this topic,
though it may just be that libgeotiff, GDAL and various other packages
need to be fixed.  I am worried if alot of packages get it wrong, and no
one gets it right (other than OSSIM so far) that we might cause more
problems by changing the common convention.

   http://bugzilla.remotesensing.org/show_bug.cgi?id=1094

Best regards,
-- 
---------------------------------------+--------------------------------
---------------------------------------+------
I set the clouds in motion - turn up   | Frank Warmerdam,
warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent


_______________________________________________
Geotiff mailing list
Geotiff at lists.maptools.org
http://lists.maptools.org/mailman/listinfo/geotiff