[Proj] Re: projection of Madagascar

Clovis clovis.grinand at ird.fr
Wed Apr 30 01:33:09 PDT 2008


Tanks for the details and for the quick answers,

Indeed laborde projection is not implemented by default in the esri 
projection library.
I retrieved this laborde projection approximation from friend working in 
a GIS company for long times. I guess is the the one that is used by 
most governement institution and companies in the country.
This approximation does produce satisfactory results (good match with 
various source of dataset). But i can't say if there decimeter offset as 
I'm not working at that scale.
Hope to see the laborde be implemented in Esri product and all other 
software.

 From the various comments, should i conclude that there is still no 
accurate laborde projection definition and transformation?
Does the 29701 or 29702 should be used as the most correct laborde 
approximation? What is the best one: 29701 or 29702?

Clovis Grinand



Melita Kennedy a écrit :
> Dear Clovis,
>
> That ArcGIS definition is an approximation; it's not the true Laborde 
> implementation and is not guaranteed to give correct answers. 
>
> We do plan to try to implement Laborde in the future, but it just 
> hasn't made it to the top of the list yet. 
>
> Melita Kennedy 
> ESRI Product Specialist
>
>   
>> Date: Tue, 29 Apr 2008 19:27:48 +0300
>> From: Clovis <clovis.grinand at ird.fr>
>> Subject: [Proj] projection of Madagascar
>> To: proj at lists.maptools.org
>>
>> Dear list,
>>
>> I have been quickly through the archive but didn't find any topic on this.
>> I have a projection issue concerning Madagascar.
>>
>> Historically there is one dedicated projection called "Laborde", using 
>> International 1925 ellipsoïd with others parameters.
>> From what i know it has been approximate with an oblique mercator 
>> projection and another called hotine oblique mercator.
>>
>> Then, I use ArcGIS usually for my application and want slowly to move 
>> towards R and QGIS, espacially to transform point coordinates in R 
>> (spTransform). The big problem is that none of three have the same 
>> projection definition and do display large offset when projecting GPS in 
>> WGS84 to "Laborde".
>>
>> From the projection definition, you will see below that there is some 
>> agreement (lat of origine, scale factor, false easting, false northing, 
>> azymuth) but some are completly different : longitude of origine and 
>> prime meridian etc.
>>
>> As I consider ArcGIS definition and transformation to be the right one 
>> (works well!), my question are:
>> - Is the EPSG parameters up to date?
>> - How can i tune the parameters to obtain the same coordinates that with 
>> arcGIS.
>> - Why lonc and not lon_0 ? (lon_0 does not give anything good)
>>
>> I hope I m not missing something obvious.
>>
>> Clovis
>>
>> # EPSG 29700 Via R
>> print(CRSargs(CRS("+init=epsg:29700")))
>> [1] " +init=epsg:29700 +proj=omerc +lat_0=-18.9 +lonc=44.10000000000001 
>> +alpha=18.9 +k=0.9995000000000001 +x_0=400000 +y_0=800000 +ellps=intl 
>> +towgs84=-189,-242,-91,0,0,0,0 +pm=paris +units=m +no_defs"
>>
>> # EPSG 29700 Via GQIS
>> +proj=omerc +lat_0=-18.9 +lonc=41.76277083000001 +alpha=18.9 
>> +k=0.9995000000000001 +x_0=400000 +y_0=800000 +ellps=intl 
>> +towgs84=-189,-242,-91,0,0,0,0 +pm=paris +units=m +no_defs
>>
>> # Laborde projection in ArcGIS
>> laborde
>> Projection: Hotine_Oblique_Mercator_Azimuth_Center
>> False_Easting: 400000,000000
>> False_Northing: 800000,000000
>> Scale_Factor: 0,999500
>> Azimuth: 18,900000
>> Longitude_Of_Center: 46,437229
>> Latitude_Of_Center: -18,900000
>>
>> GCS_Tananarive_1925
>> Name: GCS_Tananarive_1925
>> Remarks:
>> Angular Unit: Degree (0,017453292519943299)
>> Prime Meridian: Greenwich (0,000000000000000000)
>> Datum: D_Tananarive_1925
>>  Spheroid: International_1924
>>    Semimajor Axis: 6378388,000000000000000000
>>    Semiminor Axis: 6356911,946127946500000000
>>    Inverse Flattening: 297,000000000000000000
>>     
>
>
>
> _______________________________________________
> Proj mailing list
> Proj at lists.maptools.org
> http://lists.maptools.org/mailman/listinfo/proj
>
>   




More information about the Proj mailing list