gis-l: Projection File Issue

[Clay, Linc]

GIS Users,

Please be aware of the projection issue described below.  We would =
recommend
that you carefully examine the .prj files on your machine to ensure that =
you
do not have the mal-formed prj file.


Projection File Issue

A short while ago, GeoPlan discovered an inconsistency with some of our
layers' metadata.  Upon researching the metadata inconsistency, they and =
we
have discovered that ArcView 3.2a created this inconsistent projection =
file,
which was then propagated in ArcGIS for a short time.  There was no =
impact in
ArcView 3.2a.  However, when the inconsistent projection file was used =
with
ArcGIS, the data was regarded as NAD83 rather than HARN.  For a complete
description of the issue, it's lineage, and the testing we have done, =
read
the paper at the link below.

\\bis-4cgxh11\GISServer\GISSection\Projects\CY2005\Projection_Revisit\pro=
ject
ion_phaseI_appendix.doc

Although most of our layers are fine, a few of them will get new =
projection
definitions.  We will also take this opportunity to refresh some of our
third-party layers that are getting stale.  The good news is that there =
are
very few layers where the scale is such that the projection definition =
has an
impact and where the definition was inconsistent.  The other good news =
is
that we are talking differences typically of inches, and maximally of =
two
feet or so (which are only significant at data acquisition scales of =
1:12,000
or larger).

*** We will be rectifying this problem starting Friday, April 1 at 6:00 =
PM
EST.  The work should be completed by Monday, April 4 at 7:00 AM EDT.  =
While
ArcSDE will be available during most of this weekend, there may be brief
periods when it is not available. ***

How does this impact you?  We would recommend that you examine all your =
*.prj
files, any ArcIMS map services that use the COORDSYS, FEATURECOORDSYS, =
or
FILTERCOORDSYS tags, any scripts or programs that use the well-known =
text
from the *.prj file, and any ArcSDE layers in your schema for this
inconsistent projection definition.  We would recommend destroying the
inconsistent *.prj files and replacing them with a consistent *.prj file =
or
reprocessing the data where needed.

The inconsistent *.prj definition includes the following two elements.  =
Note
that the GCS and the DATUM do not agree: one is HARN-based and the other =
is
NAD83-based.

	//Inconsistent */prj file - do not use
	GEOGCS["GCS_North_American_1983_HARN"
	DATUM["D_North_American_1983"

A consistent *.prj file includes the following elements.  Note that they =
both
reference HARN.

	//Consistent *prj file
	GEOGCS["GCS_North_American_1983_HARN"
	DATUM["D_North_American_1983_HARN"

We have created consistent *.prj files for reference for both NAD83 and =
HARN.
These are located at the link below.

\\bis-4cgxh11\GISServer\GISSection\Projects\CY2005\Projection_Revisit\fde=
palb
ersharn.prj
\\bis-4cgxh11\GISServer\GISSection\Projects\CY2005\Projection_Revisit\fde=
palb
ersnad83.prj

-Linc

Linc Clay
FDEP/DRAM/BIS/GIS
2600 Blair Stone Rd, MS 6520
Tallahassee, FL 32399-2400
Voice: 850-245-8295, SC 205-8295
Fax: 850-245-8263, SC 205-8263
Email: linc.clay@dep.state.fl.us