[MetaCRS] Coordinate offset after transformation from 4267 to 4326

Martin Desruisseaux martin.desruisseaux at geomatys.com
Tue May 24 17:51:20 PDT 2016


Hello Betsy

The difference that you get may depend on whether a datum shift has been
applied or not. If both Proj.4 and GeoTools have applied a datum shift,
then it may depend on the values in the TOWGS84 expression. There is
many different possible TOWGS84 values for a transformations between
NAD27 and WGS84, depending on the geographic area. If you are
transforming coordinates in USA, you should have TOWGS84[-8, 160, 176]
(for an error below 10 metres). If you are transforming coordinates in
Canada, you should have TOWGS84[-10, 158, 187] (for an error below 20
metres). The TOWGS84 parameters in the GeoTools CRS provided with your
email are for Cuba (operation code EPSG:15978) except the sign of 0.526
which should be -0.526.

In order to compare with Proj.4, we would need to know which parameters
Proj.4 has selected. It may also use the NADCON grids in USA, or NTv2
grids in Canada. I do not know how to get this information from Proj.4.

Alternatively you could also try Apache SIS 0.7 for coordinate
transformations. It takes in account the geographic area of the points
to transform, support GML and WKT version 1 and 2, tell you which
parameters it selected, in which geographic area they are valid and what
accuracy to expect, etc. Version 0.7 should be released this week.

    Martin


Le 24/05/16 à 18:11, Betsy Emmons a écrit :
> Hello, 
>
> I'm using Proj.4 v4.9.1 and GeoTools v13.x. When a WMS layer is
> brought into QGIS as EPSG:4326 there is a noticeable offset compared
> to the same local layer. The WMS layer source is EPSG:4267. 
>
> A WMS 4326 sample point has a coordinate of -97.2341220, 28.0611696
> and the same local 4326 sample point has a coordinate of -97.2339531,
> 28.0608752. This is roughly a 40m difference. Some local calculations
> have shown the error margin to be <=111m.
>
> Here is the source 4267 projection
> definition: GEOGCS["GCS_North_American_1927",DATUM["D_North_American_1927",SPHEROID["Clarke_1866",6378206.4,294.9786982]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]]. 
>
> Here's what geotools is using:
> GEOGCS["NAD27", 
>   DATUM["North American Datum 1927", 
>     SPHEROID["Clarke 1866", 6378206.4, 294.9786982138982,
> AUTHORITY["EPSG","7008"]], 
>     TOWGS84[2.478, 149.752, 197.726, 0.526, -0.498, 0.501, 0.685], 
>     AUTHORITY["EPSG","6267"]], 
>   PRIMEM["Greenwich", 0.0, AUTHORITY["EPSG","8901"]], 
>   UNIT["degree", 0.017453292519943295], 
>   AXIS["Geodetic longitude", EAST], 
>   AXIS["Geodetic latitude", NORTH], 
>   AUTHORITY["EPSG","4267"]]
>
> Proj.4 is really a great library! Thank you for maintaining and
> contributing to it! 
>
> Thank you, 
> Betsy Emmons



More information about the MetaCRS mailing list