[Proj] ESRI's standard_parallel_1 for Mercator morphedto latitude_of_origin

[Melita Kennedy]

(yes, I know I'm top-posting and that it's not customary for the mailing list)

This may only get more confusing with the new EPSG Mercator and Stereographic variants. Currently, there are 4 Mercator variants (ignoring false easting/northing and easting/northing at false origin):

Mercator (Spherical) 

  Latitude of natural origin   
  Longitude of natural origin 

Mercator (variant A) aka 1SP

  Latitude of natural origin (must equal zero)   
  Longitude of natural origin   
  Scale factor at natural origin   

Mercator (variant B)

  Latitude of 1st standard parallel    
  Longitude of natural origin  

Mercator (variant C)

  Latitude of 1st standard parallel   
  Longitude of natural origin   
  Latitude of false origin   

Esri currently supports variant B and the Spherical version indirectly through Mercator (Auxiliary Sphere) or by using a sphere-based geographic CRS. Esri wants "Central Meridian" and "Standard Parallel 1". Internally, we use an array to store parameters. SP1 and LatOfOrigin use different slots so it wouldn't work by chance either.

Melita


>-----Original Message-----
>From: proj-bounces at lists.maptools.org
>[mailto:proj-bounces at lists.maptools.org] On Behalf Of Mikael Rittri
>Sent: Friday, October 12, 2012 2:33 PM
>To: PROJ.4 and general Projections Discussions
>Subject: Re: [Proj] ESRI's standard_parallel_1 for Mercator morphedto
>latitude_of_origin?
>
>Joe,
>by "parameter name munging", do you mean in addition to what is done by
>morphToESRI() and morphFromESRI() in OGRSpatialReference? Can you give
>examples?
>    Would you agree that for my example, morphFromESRI() makes a
>hypercorrection (trying to fix what ain't broken)?
>     By the way, I have found the analogous ticket
>http://trac.osgeo.org/gdal/ticket/3036
>for Equidistant Cylindrical. 
>
>Mikael Rittri
>Carmenta
>Sweden
>http://www.carmenta.com
>
>12 okt 2012 kl. 18:41 skrev "Joseph White" <JWhite at terragotech.com>:
>
>> Mikael,
>> It's not just you.  In the course of our software, we have to 
>> transition data back and forth from ESRI to GDAL/OGR, and we have to 
>> do some parameter name munging to get ESRI to take some of the 
>> coordinate system data.
>> 
>> Joe
>> 
>> -----Original Message-----
>> From: proj-bounces at lists.maptools.org
>> [mailto:proj-bounces at lists.maptools.org] On Behalf Of Mikael Rittri
>> Sent: Thursday, October 11, 2012 4:43 PM
>> To: PROJ.4 and general Projections Discussions
>> Subject: [Proj] ESRI's standard_parallel_1 for Mercator morphed to 
>> latitude_of_origin?
>> 
>> This question really concerns GDAL/OGR rather than Proj.4, but anyway:
>> 
>> I have had a problem that I think was caused by
>> 
>> http://trac.osgeo.org/gdal/changeset/7542
>> 
>> which caused the Mercator parameter "standard_parallel_1" in ESRI WKT 
>> to be renamed to "latitude_of_origin" during conversion to OGC WKT.
>> 
>> This makes no sense to me. These are different parameters with 
>> different semantics, and according to 
>> http://www.remotesensing.org/geotiff/proj_list/mercator_2sp.html
>> both of them are valid parameter names for Mercator (2SP) in OGC WKT.
>> 
>> So why the renaming? Is the semantics subtly different in the ESRI 
>> world?
>> Or is this just a bug that has survived because a Mercator seldom has 
>> a non-zero standard parallel?
>> 
>> I'd be grateful for comments. 
>> 
>> Mikael Rittri
>> Carmenta
>> Sweden
>> http://www.carmenta.com
>> _______________________________________________>
>_______________________________________________
>Proj mailing list
>Proj at lists.maptools.org
>http://lists.maptools.org/mailman/listinfo/proj