[Proj] looking for the right params for RSO

[Mikael Rittri]

I think the gamma parameter for omerc isn't yet supported in Proj.4.

Although the PROJ.4 timeline says it was added in March 1, 2010 
( http://trac.osgeo.org/proj/timeline?from=2010-03-17T15%3A18%3A06Z&precision=second ),
when I tried FWTools 2.4.7 for Windows, the gamma had no effect, and
the -v option (verbose) said it was unrecognized.  

Hermann Peifer gave the solution to use the +rot_conv flag instead.   
In general, the semantics of the two obscurely-named omerc flags is 
as follows, I think:

+no_uoff omitted:

    The origin, the point with projected coordinates (FalseEasting, FalseNorthing), 
    is at the projection center, on the longitude of center and latitude of center.

+no_uoff included: 

    The origin is the so-called natural origin, on the initial line near the equator.

+rot_conv omitted:

    The projected grid gets a final rotation so that grid north equals true north 
    on the projection center.  Use this when alpha (azimuth) and gamma 
    (rectified-to-grid angle) are equal, which they usually are outside Malaysia.

+rot_conv included:

    The final rotation makes grid north equal true north at the natural origin.
    Use this when alpha and gamma differs, which only happens in Malaysia. 

---

As for the datum shift, I think you can use the datum shift 
    
    EPSG:8659, "Kertau (RSO) to WGS 84 (1)"

which EPSG presents as a concatenated datum shift via "Kertau 1968". But I think 
it is equivalent to having 

    +towgs84=-11,851,5

provided you modify the equatorial radius to be +a=6377295.664 instead. 

I also noticed that if the CRS is meant to be EPSG:3168, "Kertau (RSO) / RSO Malaya (m)", 
then the False Easting of your Well-Known Text differs by 1.05 meters from the 
EPSG value, which is 804670.24 m.  (On the other hand, EPSG says that the 
accuracy of the datum shift is only 15 meters).  

Best regards,
  Mikael Rittri
  Carmenta AB
  Sweden
  www.carmenta.com

-----Original Message-----
From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of Hermann Peifer
Sent: den 29 augusti 2010 18:32
To: PROJ.4 and general Projections Discussions
Cc: Christoph Dohmen
Subject: Re: [Proj] looking for the right params for RSO

On 27/08/2010 08:57, Christoph Dohmen wrote:
> Hi list,
>
> I'm searching for the right parameter set for RSO projected coordinates.
>
> Given a dedicated point from google earth with lat-long = 100.563743E 
> 5.389524N
>
> Having a raster dataset with a projection set as the following:
>
> PROJCS["Kertau_RSO_Malaya_Meters",
>      GEOGCS["GCS_Kertau",
>          DATUM["D_Kertau",
>              SPHEROID["Everest_1830_Modified",6377304.063,300.8017]],
>          PRIMEM["Greenwich",0.0],
>          UNIT["Degree",0.0174532925199433]],
>      PROJECTION["Rectified_Skew_Orthomorphic_Natural_Origin"],
>      PARAMETER["False_Easting",804671.299775],
>      PARAMETER["False_Northing",0.0],
>      PARAMETER["Scale_Factor",0.99984],
>      PARAMETER["Azimuth",-36.97420943711801],
>      PARAMETER["Longitude_Of_Center",102.25],
>      PARAMETER["Latitude_Of_Center",4.0],
>      PARAMETER["XY_Plane_Rotation",-36.86989764584402],
>      UNIT["Meter",1.0]]
>
> Inside the dedicated point has the coordinates 286400m east 596700m north.
>
> Now I want to set up a translation first via proj directly and later 
> on for a gdal-warp to transform the rasterdataset.
>
> I tried out the parameter:
>
> cs2cs +init=epsg:4326 +to +proj=omerc +a=6377304.063 +rf=300.8017 
> +lat_0=4.0 +lonc=102.25 +alpha=323.025790562881987 
> +gamma=323.13010235415598 +k_0=0.99984 +x_0=804671.299775
>
> but the result is 617824.69 153846.36
>

You are getting somewhat closer to the expected result with:

cs2cs -v +init=epsg:4326 +to +proj=omerc +a=6377304.063 +rf=300.8017 
+lat_0=4.0 +lonc=102.25 +alpha=323.025790562881987 +k_0=0.99984
+x_0=804671.299775 +no_uoff +rot_conv

286289.13  596607.56 0.00

You might have to look for appropriate towgs84 parameter values.

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