A while back I was wondering if ArcGIS and OGR would produce the same
output from the same projection operation... Results where identical
for operations within the same datum, using a GCS -> UTM projection.
A couple days ago I did another experiment, this time incorporating a
datum shift: GCS NAD83 --> UTM NAD27. I have summarized the results
here:
http://casoilresource.lawr.ucdavis.edu/drupal/node/259
of note:
- the default tranform in ArcGIS does not produce the correct results.
- the 'NADCON' transform in ArcGIS produces the same results as OGR.
Any comments on some of the questions posed are welcomed.
Cheers,
Dylan
Hi Frank,
Yes. The only discrepancy is with the default, and oddly named,
'NAD_1927_To_NAD_1983_6' "transform" present in ArcGIS 9.0. The NADCON
approach in ArcGIS matches the output from OGR which matches the
output from the NGS NADCON webpage.
Thanks for your hard work!
Dylan
On 7/29/06, Frank Warmerdam <warmerdam@pobox.com> wrote:
Dylan Beaudette wrote:
> A while back I was wondering if ArcGIS and OGR would produce the same
> output from the same projection operation... Results where identical
> for operations within the same datum, using a GCS -> UTM projection.
>
> A couple days ago I did another experiment, this time incorporating a
> datum shift: GCS NAD83 --> UTM NAD27. I have summarized the results
> here:
>
> http://casoilresource.lawr.ucdavis.edu/drupal/node/259
>
> of note:
>
> - the default tranform in ArcGIS does not produce the correct results.
> - the 'NADCON' transform in ArcGIS produces the same results as OGR.
>
> Any comments on some of the questions posed are welcomed.
Dylan,
I read over your post quickly, and my understanding is that the outstanding
question relates to the different methods in ArcGIS, is that right? And
that OGR (and cs2cs) match the results of using ArcGIS with the NADCON
method? I'm just making sure there isn't an issue I need to be looking
into.
Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up | Frank Warmerdam, warmerdam@pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush | President OSGF, http://osgeo.org
Hi Dylan,
A couple days ago I did another experiment, this time incorporating a
datum shift: GCS NAD83 --> UTM NAD27. I have summarized the results
here: http://casoilresource.lawr.ucdavis.edu/drupal/node/259
of note:
- the default tranform in ArcGIS does not produce the correct results.
- the 'NADCON' transform in ArcGIS produces the same results as OGR.
I asked ESRI-L about this, and the near unanimous response is: there is no default transformation in ArcGIS, but an alphabetised list.
"The 'default' transformation that Dylan refers to, NAD_1927_To_NAD_1983_6, is just the first in the alphabetical list. As it turns it out, it uses NTv2 and is designed for Quebec. ArcGIS generally doesn't set any default transformations but leaves it to you to decide which one is most appropriate. Dylan's data is outside the area covered by the NTv2 grid file, so essentially no tranformation occurs when he uses NAD_1927_To_NAD_1983_6. A better reference for him, which does list the methods, parameters and areas of use for the transformations is:
http://support.esri.com/index.cfm?fa=knowledgebase.techarticles.articleShow&d=21327 "
-- Melita Kennedy Product Specialist ESRI, Inc.
Even if mistaken about the "default", thank you for conducting the experiment and telling us about it. Otherwise I would sailed along quite happily usiing the wrong selections myself. Also thanks to you we also know that provided the appropriate transformation method is chosen equivalent results can be achieved with both ArcGIS and GDAL, so we are all free to choose the tool most readily at hand.
cheers,
--
matt wilkie
--------------------------------------------
Geographic Information,
Information Management and Technology,
Yukon Department of Environment
10 Burns Road * Whitehorse, Yukon * Y1A 4Y9
867-667-8133 Tel * 867-393-7003 Fax
http://environmentyukon.gov.yk.ca/geomatics/
--------------------------------------------
Hi Matt,
On Wednesday 02 August 2006 10:24, Matt Wilkie wrote:
Hi Dylan,
> A couple days ago I did another experiment, this time incorporating a
> datum shift: GCS NAD83 --> UTM NAD27. I have summarized the results
> here: http://casoilresource.lawr.ucdavis.edu/drupal/node/259
>
> of note:
>
> - the default tranform in ArcGIS does not produce the correct results.
> - the 'NADCON' transform in ArcGIS produces the same results as OGR.
I asked ESRI-L about this, and the near unanimous response is: there is
no default transformation in ArcGIS, but an alphabetised list.
"The 'default' transformation that Dylan refers to,
NAD_1927_To_NAD_1983_6, is just the first in the alphabetical list. As
it turns it out, it uses NTv2 and is designed for Quebec. ArcGIS
generally doesn't set any default transformations but leaves it to you
to decide which one is most appropriate. Dylan's data is outside the
area covered by the NTv2 grid file, so essentially no tranformation
occurs when he uses NAD_1927_To_NAD_1983_6. A better reference for him,
which does list the methods, parameters and areas of use for the
transformations is:
http://support.esri.com/index.cfm?fa=knowledgebase.techarticles.articleShow
&d=21327 "
-- Melita Kennedy Product Specialist ESRI, Inc.
Ok. I didn't even notice that there was some order in the list. Then I stand
corrected, in that this was not a 'default' transform, just the 'first' one
in the list. Seems reasonable.
Even if mistaken about the "default", thank you for conducting the
experiment and telling us about it. Otherwise I would sailed along quite
happily usiing the wrong selections myself. Also thanks to you we also
know that provided the appropriate transformation method is chosen
equivalent results can be achieved with both ArcGIS and GDAL, so we are
all free to choose the tool most readily at hand.
cheers,
Sure. Thanks for taking time to do some of the detective work. I have added a
small blub at the beginning of the original post, so that future visitors
will not have to pour over the comments to get at the final conclusions.
http://casoilresource.lawr.ucdavis.edu/drupal/node/259
Cheers,
--
Dylan Beaudette
Soils and Biogeochemistry Graduate Group
University of California at Davis
530.754.7341
Hi Dylan,
Sure. Thanks for taking time to do some of the detective work.
You're welcome. People have done the same for me a number of times now so I figured I'd best return the favour. 
take care,
--
matt wilkie
--------------------------------------------
Geographic Information,
Information Management and Technology,
Yukon Department of Environment
10 Burns Road * Whitehorse, Yukon * Y1A 4Y9
867-667-8133 Tel * 867-393-7003 Fax
http://environmentyukon.gov.yk.ca/geomatics/
--------------------------------------------