[GeoNode-users] Layers with epsg:4618 shiffting in GeoNode

Daniel Victoria daniel.victoria at gmail.com
Tue Jun 16 13:19:09 PDT 2015


Thanks all. I believe I found the problem, thanks to all your help.

So, the strange thing is the towgs84 values assined to EPSG:4618.
Gdal/OGR is usingwhatever comes from proj, which is
<4618> +proj=longlat +ellps=aust_SA +towgs84=-57,1,-41,0,0,0,0 +no_defs  <>

But GeoServer uses

TOWGS84[-66.87, 4.37, -38.52, 0.0, 0.0, 0.0, 0.0],

Which, according to the URL sent by Stefan, is the TOWGS84 used by EPSG:4291.

Actually, the URL about Brazilian Coordinate systems sent by Stefan
says that there are some problems regarding the towgs84 values in GDAL
http://wiki.osgeo.org/wiki/Brazilian_Coordinate_Reference_Systems#Summary_Table


But since GeoNode says the uploaded data is EPSG:4326, I believe
GeoServer is not understanding the SAD69 projection info.


So, the take home message is that SAD69 definition is a bag of mixed
nuts... Better stick with something with better definitions. Meanwhile,
I'll try to get the correct towgs84 values and send to someone in the proj
list (or gdal or who knows).

Bellow are the projection data for the shapefile, ogrinfo etc etc if anyone
woul'd like to check

Cheers and thanks
Daniel

This is the shape .prj file
GEOGCS["SAD69", DATUM["D_South_American_1969",
SPHEROID["GRS_1967_Truncated",6378160,298.25]],
PRIMEM["Greenwich",0],
UNIT["Degree",0.017453292519943295]]

This is how QGis sees in the metadata tab
+proj=longlat +ellps=aust_SA +towgs84=-57,1,-41,0,0,0,0 +no_defs

This is what ogrinfo sees:
INFO: Open of `erros_00167.shp'
      using driver `ESRI Shapefile' successful.

Layer name: erros_00167
Geometry: 3D Point
Feature Count: 235
Extent: (-67.591170, -10.430618) - (-66.788872, -9.600199)
Layer SRS WKT:
GEOGCS["SAD69",
    DATUM["South_American_Datum_1969",
        SPHEROID["GRS_1967_Modified",6378160,298.25]],
    PRIMEM["Greenwich",0],
    UNIT["Degree",0.017453292519943295]]
FEAT_ID: Integer (10.0)
ERROR: String (254.0)

This is what spatialreference.org gives for EPSG:4618
+proj=longlat +ellps=aust_SA +no_defs
http://spatialreference.org/ref/epsg/4618/proj4/

Abd this is the what Geoserver thinks of 4618

GEOGCS["SAD69",
  DATUM["South American Datum 1969",
    SPHEROID["GRS 1967 Modified", 6378160.0, 298.25, AUTHORITY["EPSG","7050"]],
    TOWGS84[-66.87, 4.37, -38.52, 0.0, 0.0, 0.0, 0.0],
    AUTHORITY["EPSG","6618"]],
  PRIMEM["Greenwich", 0.0, AUTHORITY["EPSG","8901"]],
  UNIT["degree", 0.017453292519943295],
  AXIS["Geodetic longitude", EAST],
  AXIS["Geodetic latitude", NORTH],
  AUTHORITY["EPSG","4618"]]





On Tue, Jun 16, 2015 at 4:27 PM, Simone Dalmasso <simone.dalmasso at gmail.com>
wrote:

> That's correct, GeoNode is not reprojecting, the issue should be somewhere
> in geoserver.
>
> 2015-06-16 21:16 GMT+02:00 Stefan Steiniger <sstein at geo.uzh.ch>:
>
>>  Hi Daniel (thanks for re-routing back to the list)
>>
>> If it is indeed the projection, then the problem is with GeoServer and
>> the projection library used (geotools?).
>> So, I doubt that parsing the code is the problem, but the parameters set
>> in GeoServer, and not with GeoNode.
>>
>> see this too:
>>
>> http://docs.geoserver.org/stable/en/user/advanced/crshandling/customcrs.html
>>
>> cheers from Santiago de Chile,
>> stefan
>>
>>
>> On 06/16/2015 03:59 PM, Daniel Victoria wrote:
>>
>>    -- replying to the list --
>>
>>  Hi Stefan,
>>
>>  Thanks for the input. EPSG:4618 is a geographic coordinate system based
>> on the South American Datum. And apparently somewhere GeoNode is not liking
>> this coordinate system. If I convert the data to EPSG:4326 in QGis prior to
>> uploading, it all works fine. Could it be that QGIS understands the .prj
>> file for the EPSG:4618 but GeoNode does not and assumes EPSG:4326?
>>
>>  By looking at the ISO metadata from the EPSG:4618 layer, it tells me
>> that the data is in EPSG:4326. Does GeoNode converts the SRS on layer
>> upload? Or is that telling me that my projection file was bogus?
>>
>>  Thanks
>>  Daniel
>>
>>
>>
>> On Tue, Jun 16, 2015 at 2:29 PM, Stefan Steiniger <sstein at geo.uzh.ch>
>> wrote:
>>
>>>  perhaps a silly question:
>>>
>>> (a) are you indeed having the same coordinate system (same prj files?)
>>> for both shape files?
>>> (b) what happens if you not use epsg:4618 but some UTM? (sorry can't
>>> figure exactly what 4618 is... it looks like geographic coordinates, so
>>> perhaps a projection makes sense)
>>>
>>> see also this:
>>> http://wiki.osgeo.org/wiki/Brazilian_Coordinate_Reference_Systems
>>>
>>> cheers,
>>> stefan
>>>
>>>
>>> On 06/16/2015 11:26 AM, Daniel Victoria wrote:
>>>
>>>>>>  point_error.7z
>>> <https://drive.google.com/file/d/0B5VE16NDdlfERE5yenZNaHY2bTA/edit?usp=drive_web>
>>> ​​
>>>  TC_AC_2010_00167.zip
>>> <https://drive.google.com/file/d/0B5VE16NDdlfEaGNHZ2ZSU0RjMHM/edit?usp=drive_web>
>>> ​I've got two shapefiles with epsg:4618. One is a polygon layer with
>>> loads of geometry errors and the second is a point layer with the error
>>> location. When I load both in QGIS, polygons and errors are aligned
>>> (layers_qgis.png). I then uploaded the layers to GeoNode (2.4b24) and
>>> points and polygons have shifted (layers_geonode.png).
>>>
>>> I then added back the two layers in QGis using WMS services. The points
>>> layer shifted a little (~8 meters - point_error.png) but polygons shifted
>>> 69 m (poly_error.png).
>>>
>>> Any idea what could be wrong?
>>>  The shapefiles (poligon and points) can be downloaded from the google
>>> drive links
>>>
>>>  Thanks
>>>  Daniel
>>>
>>>
>>>  _______________________________________________
>>> geonode-users mailing listgeonode-users at lists.osgeo.orghttp://lists.osgeo.org/cgi-bin/mailman/listinfo/geonode-users
>>>
>>>
>>>
>>
>>
>> _______________________________________________
>> geonode-users mailing list
>> geonode-users at lists.osgeo.org
>> http://lists.osgeo.org/cgi-bin/mailman/listinfo/geonode-users
>>
>>
>
>
> --
> Simone
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/geonode-users/attachments/20150616/12edd6be/attachment.html>


More information about the geonode-users mailing list