[geotk] Transform from EPSG 4979 to 3D CRS

[Aaron Braeckel]

Are there any authorities (ex: EPSG) that define a general-purpose
ellipsoidal vertical CRS analogous to EPSG:5714?  I'd rather not define
my own authority and code if one already exists and I cannot make use of
the Java classes in this case.  Your first approach would work, but I
need to be able to externally refer to an independent ellipsoidal vertCRS.

Aaron

On 05/03/2011 05:15 AM, Martin Desruisseaux wrote:
> Hello Aaron
>
> The issue is probably there:
>
> Le 03/05/11 06:12, Aaron Braeckel a écrit :
>> (...snip...)
>> CoordinateReferenceSystem crs4979 = CRS.decode( "EPSG:4979" ); //3D
>> CRS - lat,lon,ellipsoidal height
>> (...snip...)
>> CoordinateReferenceSystem vertCRS = CRS.decode( "EPSG:5714"); //mean
>> sea level height
>> CoordinateReferenceSystem compoundCRS = new
>> DefaultCompoundCRS("compound", lambert, vertCRS );
>> CRS.findMathTransform( crs4979, compoundCRS );  //this fails with "No
>> transformation available from system "GeographicCRS[WGS 84]" to
>> "CompoundCRS[ CompoundCRS ]"."
>
> The source CRS (EPSG:4979) uses an ellipsoidal height. The target CRS
> uses a "mean sea level" height (EPSG:5714). The current Geotk version
> doesn't know by itself how to go from ellispoidal height to mean sea
> level.
>
> I see two possible workarounds:
>
>    * Use the DefaultVerticalCRS.ELLIPSOIDAL constant instead than
> EPSG:5714
>      (easiest approach).
>    * Instantiate explicitely a EarthGravitationalModel
>     
> <http://www.geotoolkit.org/apidocs/org/geotoolkit/referencing/operation/transform/EarthGravitationalModel.html>
>      transform from ellipsoidal height to geoidal height.
>
>
> There is a JIRA task for implementing automatic instantiation of
> EarthGravitationalModel: http://jira.geotoolkit.org/browse/GEOTK-139
>
> Regards,
>
>     Martin
>
> _______________________________________________
> Geotoolkit mailing list
> Geotoolkit at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/geotoolkit