Skipping to part two - spanning UTM zones.

There are a number of question you need to answer regarding your need
for spatial accuracy (how you intend to use the data).  There is not a
simple answer without knowing more of your requirements.

It is possible to store data from different coordinate systems / zones
in a single database table such as in Oracle or using SpatialWare and
also SDE.  There are several "costs" that should be considered.  
a) Whenever data is extracted the system must interrogate the coordinate
system within the geometry and if required perform a transformation.
Depending upon the client a second transformation may occur to transform
to the current map window coordinate system.  While automatic we are
talking CPU cycles.
b) Data administration is a bit more complex in that multiple coordinate
systems are contained in a single table.  The ramifications must be
considered when performing geometry editing tasks. 

The other approach to this problem is to store data in a single
coordinate system (LAT LONG WGS84 for example) and convert to the
desired UTM zone as data is extracted (In Mapinfo if you set up the map
window first with the desired coordinate system for that session then
all subsequent data will be transformed upon display).  This means
taking the source data and transforming it to a "common" coordinate
system. Of course there are consideration if you maintain the geometry
and have to perform many transformations, or editing tasks that require
higher degrees of accuracy (say third order control or better). If you
are using source data such as from MapInfo or GDT based upon 1:100,000
or 1:24000 map accuracy then this may not be an issue as your spatial
accuracy (resolution) is quite low (40' for 1:24000).  Remember any time
you transform there are rounding error consideration as well as the
transformation model accuracy, this should always be considered in
respect to the accuracy of the data collection methods.  At 1:100000
your accuracy is within 165' - what's a few feet going to mean? This
works for data where the spatial accuracy is not the driving concern.
If you need to locate a property boundary, pole location, etc, within a
few feet accuracy this won't do, and don't even consider it when using
real property descriptions.

Also note that if you select a single UTM zone as your "origin" zone
(say zone 11) as you move into adjacent zones (east or west) the
accuracy of the data is decreased due to the transformation model.
Going into the middle of one adjacent zone (say zone 12) might result in
acceptable measurement results.  But transforming data from two zones
away (say zone 13 or 14) would result in significant distortion. 

This is often a problem for local governments where they span two state
plane coordinate zones.  They are forced to maintain "source data" in
the correct zone coordinates, but will merge the two into a single set
of tables (files) by transforming the one zone into the other.  For most
purposes this will be acceptable.  But all editing is performed on the
original files / zone coordinates to maintain accuracy.



-----Original Message-----
From: Frank Aaron (TX/EUS) [mailto:[EMAIL PROTECTED] 
Sent: Tuesday, January 06, 2004 09:34 AM
To: [EMAIL PROTECTED]
Subject: MI-L GIS Data Manipulation Tool Recommendations

Hi All,

Does anyone know of a tool that would allow one to "extract" a defined
area from a "larger" Map Database (i.e., Elevation, Clutter 
and Tiger Vectors) and create a smaller one (retaining the Map
Projection, etc.)? I was also hoping that someone might be able
to recommend a tool that might be able to merge map databases for
neighboring UTM Zones into a single database (preserving
their unique map projections). TIA,

Frank Aaron, MSc. Physics, MSEE
Sr Wireless Systems Engineer, Professional Services
Ericsson USA
Global Services North America
Tel:          (972) 583-0112
Fax:         (972) 583-1807
Mobile:    (972) 679-9291
mail to:    [EMAIL PROTECTED]





---------------------------------------------------------------------
List hosting provided by Directions Magazine | www.directionsmag.com |
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
Message number: 9782

Reply via email to