[Proj] cs2cs returns error then 0s with datum conversion

[Mikael Rittri]

Clifford,
you wrote:

> NADCON is a U.S. Federal standard.  Converting its data files to NTv2 format 
> would be useless for applications within the U.S. as it would no longer be 
> compliant with the Federal Register mandate.  
> [...]
> When working with National Coordinate Systems and Reference Frames, 
> one can get into trouble quickly through ignorance of local laws.

Right, I should have added "this suggestion is void where prohibited by law". 

Well, seriously, I would have agreed with you if we had discussed Great Britain
and its national grid shift file, for two reasons:

1) Great Britain has redefined its OSGB 1936 datum, so that
   their grid shift file is now exactly right by definition. 
   http://www.ordnancesurvey.co.uk/oswebsite/gps/docs/Geomatics_world.pdf

2) The British grid shift file is rectangular in a projected plane, 
   while NTv2 is rectangular in Lat/Long, so you cannot get millimeter
   agreement, unless you used a much denser NTv2 grid, perhaps. 
   (I have tested Paul Kelly's NTv2 version for Great Britain, 
    see http://lists.maptools.org/pipermail/proj/2006-November/002658.html , 
    and the errors were usually a couple of centimeters, if I remember right.)

But it's different for the U.S., isn't it?  The NADCON does not redefine 
the meaning of NAD27, so any software-based datum shift will be slightly 
wrong anyway.  And NTv2 and NADCON are both rectangular in Lat/Long
and use the same bilinear interpolation technique, so I expect
that you can get sub-millimeter agreement between their results.
(Maybe Frank knows whether this is true.) 

As for NADCON being a Federal Standard, I found the following notice:
 
http://www.ngs.noaa.gov/PUBS_LIB/FedRegister/FRdoc90-18809.pdf

but it says

    "It is not the intent of the notice to declare when to 
     use a datum transformation or by what method, but only 
     to declare that when a mathematical transformation is
     appropriate, NADCON is recommended."

As I understand this, other mathematical transformations (or 
other software) are still legal, though not recommended. And 
if an NTv2 version would give sub-millimeter agreement with
NADCON, I would say that it would follow the spirit of the 
recommendation, anyway. 

Just my 20 öre,

Mikael Rittri
Carmenta
Sweden
http://www.carmenta.com

________________________________

From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of Clifford J Mugnier
Sent: Wednesday, January 12, 2011 9:05 PM
To: PROJ.4 and general Projections Discussions
Subject: Re: [Proj] cs2cs returns error then 0s with datum conversion


NADCON is a U.S. Federal standard.  Converting its data files to NTv2 format would be useless for applications within the U.S. as it would no longer be compliant with the Federal Register mandate.  
 
The converse would likely apply to some other countries, including Canada and the Union of South Africa.  Other countries may have not (yet) enacted legislation that codifies the specific methodology, but I have never seen a study of such a topic.
 
When working with National Coordinate Systems and Reference Frames, one can get into trouble quickly through ignorance of local laws.
 
Clifford J. Mugnier, C.P., C.M.S.
Chief of Geodesy,
Center for GeoInformatics
Department of Civil Engineering 
Patrick F. Taylor Hall 3223A
LOUISIANA STATE UNIVERSITY 
Baton Rouge, LA  70803
Voice and Facsimile:  (225) 578-8536 [Academic] 
Voice and Facsimile:  (225) 578-4578 [Research] 
Cell: (225) 328-8975 [Academic & Research]
Honorary Life Member of the 
Louisiana Society of Professional Surveyors 
Fellow Emeritus of the ASPRS 
Member of the Americas Petroleum Survey Group


________________________________

From: proj-bounces at lists.maptools.org on behalf of Frank Warmerdam
Sent: Wed 12-Jan-11 13:25
To: Mikael Rittri
Cc: PROJ.4 and general Projections Discussions
Subject: Re: [Proj] cs2cs returns error then 0s with datum conversion



On 11-01-12 04:31 AM, Mikael Rittri wrote:
> Eric wrote:
>
>> Maybe someday an enterprising soul will convert cs2cs to use National Geodetic Survey's
>> current correction file format directly...
>
> Or maybe Frank would be willing to convert these files (conus, alaska, etc.)
> to the NTv2 format (and publish them), since he has already developed tools
> for that conversion, and since cs2cs already understands that format.
>
> (See http://lists.maptools.org/pipermail/proj/2010-August/005330.html )
>
> Of course, this would help only if the NTv2 format is unaffected by
> the 32-bit/64-bit issue.  I don't know for certain if that's true.

Folks,

The NTv2 format is not affected by structure packing, byte order or word
size and is a good candidate for a cross platform replacement for the
binary format into which the conus and related files are currently translated.
Unfortunately my work in GDAL with NTv2 format does not include support for
the .lla format yet, but I would love to replace it.  I think I have seen
the same data in .los/.las format, so perhaps I could just find and translate
those.  It would be wonderful to cleanup this aspect of proj.4 datum handling.

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

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