[Proj] Transverse Mercator algorithm with good accuracy/speedtrade-off?

[Clifford J Mugnier]

There's a lot of "other field" specialists that get into GIS.  I see a good number of people in the Oil & Gas Exploration and Production industry.  Lots of Geologists and Petroleum Engineers are found, but nary a classical Cartographer, or Geodesist in the bunch for the most part.  (Royal Dutch Shell is an exception.)  These people work large to enormous geographic areas all over the world, and don't know diddly about map projections or grids.  Most don't know the difference between UTM, Datums, and Grids muchless realize that the words are not interchangeable.  They usually are well-educated in their fields, and simply consider GIS a necessary evil tool to use in order to get their job done.  They usually don't care about important concepts in positioning tech nology, coordinate systems, reference frames, etc.  They are professionals in something other than what it is that we do, for the most part.
 
Consider a local application just for the Oil Patch:  The Gulf of Mexico.  Platforms are offshore in Louisiana and Texas, but production is petering out and exploration and development are further out in Federal Waters.  The State Waters use the various State Plane Coordinate Systems (Lambert Conformal Conic in multiple zones), but the Federal Waters use UTM in more than one zone.  Furthermore, there's the Safety Fairways in the Gulf of Mexico that use the ellipsoidal Mercator that is overlain with all of the Lambert zones and all of the UTM zones.  Geologists could not care less.  It's pretty neat for Geodesists and the classical Cartographers that understand loxodromes and Grid intersection problems, but the average GIS practitioner doesn't know the difference muchless care.
 
They want to use "UTM" as a single zone (Zone 15) for the entire Gulf of Mexico, and they want the numbers in the coordinates to be able to repeat on themselves for the whole Gulf.  Period.  You can't find one in a hundred that knows about convergence angle or scale factor, muchless cares ... they're doing something else ... or so they say.
 
There is a real need for computational precision for such applications.  Of course there's a need for education, but that's a windmill for all intents and purposes in the Oil Patch.  When it's really important, they'll have a Land Surveyor or Geodesist or classical Cartographer do the computations, but for day to day work it's the Technicians that do the GIS applications.
 
Consider the rest of the world when you read above for merely the Gulf of Mexico.
 
Dozier's paper does have a practical application for the neophyte.  It can keep them (slightly) out of trouble if the computational precison can be maintained.
 
Cliff Mugnier
LSU
 
 
 
 
 

________________________________

From: proj-bounces at lists.maptools.org on behalf of support.mn at elisanet.fi
Sent: Sat 24-May-08 16:12
To: PROJ.4 and general Projections Discussions
Subject: Re: [Proj] Transverse Mercator algorithm with good accuracy/speedtrade-off?



> The tmerc of Proj4 is efficient. But it has millimeter accuracy only out to 6° of longitude from the central meridian...


1) The most easy way is to make sure never be futher than 6° away from the central median.

or

2) You could implement a floating central median (let's call it "ftmerc"), which would automatically shift as your longitude shifts. Let's say in steps of 1 degree or less. Not at all difficult to implement. And most of "tmerc" routines can be reused.

or

3) Are we talking about UTM...

Regards: Janne



_______________________________________________
Proj mailing list
Proj at lists.maptools.org
http://lists.maptools.org/mailman/listinfo/proj


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20080524/89942c38/attachment.html>