CRS Issue #35459

tveinot · 2020-03-30T13:51:55Z

Describe the bug
Working with EPSG 2954 all exports drop to depreciated CRS 2292

How to Reproduce
Export or add any dataset using 2954 and QGIS will assign it 2292. If using a fielgeodatabase QGIS will assign 2954 but any export to shapefile, or any other spatial format I tried, QGIS will write the CRS as 2292

QGIS and OS versions
3.10.4-A Coruña

The attached zip includes a filegeodatabase and a shapefile the shapefile is an export of the same featureclass in the filegeodatabase. You wil notice that the shapefile reads as 2292 and the fieldgeodatabase reads as 2954. When I exported I selected 2954 as the EPSG to use.

This issue extends to everything I use in QGIS in 2954 from geoprocessing outputs to adding layers; QGIS keeps reverting them to 2292.
Tanks for the help
Tyler

GIS.zip

gioman · 2020-03-30T14:09:08Z

@tveinot what operating system?

tveinot · 2020-03-30T16:07:45Z

Windows 10

gioman · 2020-03-30T16:47:32Z

This is what ogrinfo (from GDAL 3 installation that comes with QGIS 3.10/12) says about your filegeodatabase layer:

C:\Users\qgis\Desktop\GIS>ogrinfo -so Test.gdb point
INFO: Open of `Test.gdb'
      using driver `OpenFileGDB' successful.

Layer name: point
Geometry: 3D Measured Point
Feature Count: 1
Extent: (389632.467800, 689808.842500) - (389632.467800, 689808.842500)
Layer SRS WKT:
PROJCRS["NAD83(CSRS) / Prince Edward Isl. Stereographic (NAD83)",
    BASEGEOGCRS["NAD83(CSRS)",
        DATUM["NAD83 Canadian Spatial Reference System",
            ELLIPSOID["GRS 1980",6378137,298.257222101,
                LENGTHUNIT["metre",1]]],
        PRIMEM["Greenwich",0,
            ANGLEUNIT["degree",0.0174532925199433]],
        ID["EPSG",4617]],
    CONVERSION["Prince Edward Isl. Stereographic (NAD83)",
        METHOD["Oblique Stereographic",
            ID["EPSG",9809]],
        PARAMETER["Latitude of natural origin",47.25,
            ANGLEUNIT["degree",0.0174532925199433],
            ID["EPSG",8801]],
        PARAMETER["Longitude of natural origin",-63,
            ANGLEUNIT["degree",0.0174532925199433],
            ID["EPSG",8802]],
        PARAMETER["Scale factor at natural origin",0.999912,
            SCALEUNIT["unity",1],
            ID["EPSG",8805]],
        PARAMETER["False easting",400000,
            LENGTHUNIT["metre",1],
            ID["EPSG",8806]],
        PARAMETER["False northing",800000,
            LENGTHUNIT["metre",1],
            ID["EPSG",8807]]],
    CS[Cartesian,2],
        AXIS["easting (E(X))",east,
            ORDER[1],
            LENGTHUNIT["metre",1]],
        AXIS["northing (N(Y))",north,
            ORDER[2],
            LENGTHUNIT["metre",1]],
    USAGE[
        SCOPE["unknown"],
        AREA["Canada - Prince Edward Island"],
        BBOX[45.9,-64.49,47.09,-61.9]],
    ID["EPSG",2954]]
Data axis to CRS axis mapping: 1,2
FID Column = OBJECTID
Geometry Column = SHAPE

Then you can convert it to SHP and GPKG using ogr2ogr:

C:\Users\qgis\Desktop\GIS>ogr2ogr -f "ESRI shapefile" -t_srs EPSG:2954 as_shape.shp Test.gdb point

C:\Users\qgis\Desktop\GIS>ogr2ogr -f "GPKG" -t_srs EPSG:2954 as_gpkg.gpkg Test.gdb point

and check again the results with ogrinfo:

C:\Users\qgis\Desktop\GIS>ogrinfo -so as_shape.shp as_shape
INFO: Open of `as_shape.shp'
      using driver `ESRI Shapefile' successful.

Layer name: as_shape
Metadata:
  DBF_DATE_LAST_UPDATE=2020-03-30
Geometry: 3D Measured Point
Feature Count: 1
Extent: (389632.467800, 689808.842500) - (389632.467800, 689808.842500)
Layer SRS WKT:
PROJCRS["NAD83(CSRS98) / Prince Edward Isl. Stereographic (NAD83)",
    BASEGEOGCRS["NAD83(CSRS98)",
        DATUM["NAD83 Canadian Spatial Reference System",
            ELLIPSOID["GRS 1980",6378137,298.257222101,
                LENGTHUNIT["metre",1]]],
        PRIMEM["Greenwich",0,
            ANGLEUNIT["degree",0.0174532925199433]],
        ID["EPSG",4140]],
    CONVERSION["Prince Edward Isl. Stereographic (NAD83)",
        METHOD["Oblique Stereographic",
            ID["EPSG",9809]],
        PARAMETER["Latitude of natural origin",47.25,
            ANGLEUNIT["degree",0.0174532925199433],
            ID["EPSG",8801]],
        PARAMETER["Longitude of natural origin",-63,
            ANGLEUNIT["degree",0.0174532925199433],
            ID["EPSG",8802]],
        PARAMETER["Scale factor at natural origin",0.999912,
            SCALEUNIT["unity",1],
            ID["EPSG",8805]],
        PARAMETER["False easting",400000,
            LENGTHUNIT["metre",1],
            ID["EPSG",8806]],
        PARAMETER["False northing",800000,
            LENGTHUNIT["metre",1],
            ID["EPSG",8807]]],
    CS[Cartesian,2],
        AXIS["easting (E(X))",east,
            ORDER[1],
            LENGTHUNIT["metre",1]],
        AXIS["northing (N(Y))",north,
            ORDER[2],
            LENGTHUNIT["metre",1]],
    USAGE[
        SCOPE["unknown"],
        AREA["Canada - Prince Edward Island"],
        BBOX[45.9,-64.49,47.09,-61.9]],
    ID["EPSG",2292]]
Data axis to CRS axis mapping: 1,2
FID: Integer64 (11.0)

C:\Users\qgis\Desktop\GIS>ogrinfo -so as_gpkg.gpkg point
INFO: Open of `as_gpkg.gpkg'
      using driver `GPKG' successful.

Layer name: point
Geometry: 3D Measured Point
Feature Count: 1
Extent: (389632.467800, 689808.842500) - (389632.467800, 689808.842500)
Layer SRS WKT:
PROJCRS["NAD83(CSRS) / Prince Edward Isl. Stereographic (NAD83)",
    BASEGEOGCRS["NAD83(CSRS)",
        DATUM["NAD83 Canadian Spatial Reference System",
            ELLIPSOID["GRS 1980",6378137,298.257222101,
                LENGTHUNIT["metre",1]]],
        PRIMEM["Greenwich",0,
            ANGLEUNIT["degree",0.0174532925199433]],
        ID["EPSG",4617]],
    CONVERSION["Prince Edward Isl. Stereographic (NAD83)",
        METHOD["Oblique Stereographic",
            ID["EPSG",9809]],
        PARAMETER["Latitude of natural origin",47.25,
            ANGLEUNIT["degree",0.0174532925199433],
            ID["EPSG",8801]],
        PARAMETER["Longitude of natural origin",-63,
            ANGLEUNIT["degree",0.0174532925199433],
            ID["EPSG",8802]],
        PARAMETER["Scale factor at natural origin",0.999912,
            SCALEUNIT["unity",1],
            ID["EPSG",8805]],
        PARAMETER["False easting",400000,
            LENGTHUNIT["metre",1],
            ID["EPSG",8806]],
        PARAMETER["False northing",800000,
            LENGTHUNIT["metre",1],
            ID["EPSG",8807]]],
    CS[Cartesian,2],
        AXIS["easting (E(X))",east,
            ORDER[1],
            LENGTHUNIT["metre",1]],
        AXIS["northing (N(Y))",north,
            ORDER[2],
            LENGTHUNIT["metre",1]],
    USAGE[
        SCOPE["unknown"],
        AREA["Canada - Prince Edward Island"],
        BBOX[45.9,-64.49,47.09,-61.9]],
    ID["EPSG",2954]]
Data axis to CRS axis mapping: 1,2
FID Column = OBJECTID
Geometry Column = SHAPE

As you can see the CRS definitions differ in the SHP and GPKG this should be the reason of what you are observing, anyway this seems to be a GDAL issue, not a QGIS one.

roya0045 · 2020-03-30T17:05:39Z

@rouault seems that this site states otherwise https://spatialreference.org/ref/epsg/2292/ is it the site that is wrong or could the deprecated status have been inverted by mistake somewhere?

tveinot · 2020-03-30T17:18:37Z

The only difference between EPSG 2292 and EPSG 2954 is the Epoc, 2292 is Epco 98 where as 2954 is Epco 2010. 2292 is suppose to be depreciated and 2954 is the one that is the one we use on the island.
EDIT: OK, according to your link 2954 is from 2002, NR Can says 2954 is using CSRS 2010 and that 2292 is using CSRS 98. So now I am just confused on what is different.
Either way I am absolutely positive that 2954 is not depreciated and 2292 is depreciated.
I appologize for all the confusion.

jmckenna · 2020-03-30T17:21:44Z

2292 was deprecated in the 1990s, and is no longer included in the current EPSG registry. The PROJ (formerly PROJ.4) values of 2292 and 2954 are the exact same. QGIS guesses at the projection, incorrectly picking the deprecated 2292, instead of the correct 2954.

rouault · 2020-03-30T17:24:17Z

@rouault seems that this site states otherwise https://spatialreference.org/ref/epsg/2292/ is it the site that is wrong or could the deprecated status have been inverted by mistake somewhere?

Generally spatialreference.org should not be considered as authoritative, but here it is indeed true. I wrongly inverted the codes in my previous answer (deleted now to avoid any confusion). So EPSG:2292 has been deprecated by EPSG:2954. And indeed, there's an issue with PROJ. It wrongly identifies PROJCS["NAD_1983_CSRS_Prince_Edward_Island",GEOGCS["GCS_North_American_1983_CSRS",DATUM["D_North_American_1983_CSRS",SPHEROID["GRS_1980",6378137.0,298.257222101]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]],PROJECTION["Double_Stereographic"],PARAMETER["False_Easting",400000.0],PARAMETER["False_Northing",800000.0],PARAMETER["Central_Meridian",-63.0],PARAMETER["Scale_Factor",0.999912],PARAMETER["Latitude_Of_Origin",47.25],UNIT["Meter",1.0]] to deprecated EPSG:2292, instead of non-deprecated EPSG:2954 . I've filed the issue upstream as OSGeo/PROJ#2116

jmckenna · 2020-03-30T17:30:13Z

@roya0045 best source is to go right to the source: go here and click the 'search by code' tab, to read of the 2292/2954 history https://www.epsg-registry.org

tveinot · 2020-03-30T17:42:29Z

Since this went out as an issue with PROJ I have andother question; So since moving to PEI I have been using 2954 in both Esri and QGIS. Esri says it is a "Double Stereographic Projection" but QGIS says it is a "Stereographic Projection" other sites like epsg.io/2954 say it is "Stereographic" yet The Government of PEI http://www.gov.pe.ca/gis/index.php3?number=77865&lang=J calls it a "Stereographic Double Projection" University of Prince Eward Island calls it a "Double Stereographic Projection" and NR Can doesn't seem to specify. The locals and the surveyors that are using 2954 (we still have a stubbron group using NAD27 Imperial lol), they say it is indeed a Double Stereographic Projection. Should this also be addressed? How do we find out who is right?

jmckenna · 2020-03-30T17:49:33Z

@tveinot maybe your comment about PROJ should be pasted into the new PROJ ticket. (be sure also to at least click the notification button on that ticket). Also, your question could be a good discussion topic to raise on the dedicated PROJ forum, which exists for this purpose (you ask good questions ha!). https://lists.osgeo.org/mailman/listinfo/proj I personally don't know the answer.

rouault · 2020-03-30T17:52:08Z

Esri says it is a "Double Stereographic Projection" but QGIS says it is a "Stereographic Projection"

Discussed here: http://geotiff.maptools.org/proj_list/oblique_stereographic.html

tveinot · 2020-03-30T18:11:53Z

@tveinot maybe your comment about PROJ should be pasted into the new PROJ ticket. (be sure also to at least click the notification button on that ticket). Also, your question could be a good discussion topic to raise on the dedicated PROJ forum, which exists for this purpose (you ask good questions ha!). https://lists.osgeo.org/mailman/listinfo/proj I personally don't know the answer.

Thank you, I did as you suggested. To my understanding GeoNB (New Brunswick) is the Geomatics Authority PE has used to establish their CRS. I would say that they are the authority but again I am not too sure.

gioman · 2020-03-30T18:59:47Z

Closing as upstream issue.

tveinot added the Bug Either a bug report, or a bug fix. Let's hope for the latter! label Mar 30, 2020

gioman added the Projections/Transformations Related to coordinate reference systems or coordinate transformation label Mar 30, 2020

gioman added the Feedback Waiting on the submitter for answers label Mar 30, 2020

rouault mentioned this issue Mar 30, 2020

Wrong identification of ESRI WKT to deprecated EPSG:2292 instead of EPSG:2954 OSGeo/PROJ#2116

Closed

gioman closed this as completed Mar 30, 2020

gioman mentioned this issue Apr 20, 2020

QGIS 3.12 not handling EPSG 2957 Always defaults to EPSG 2151 #35876

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

CRS Issue #35459

CRS Issue #35459

tveinot commented Mar 30, 2020

gioman commented Mar 30, 2020

tveinot commented Mar 30, 2020

gioman commented Mar 30, 2020

roya0045 commented Mar 30, 2020

tveinot commented Mar 30, 2020 •

edited

Loading

jmckenna commented Mar 30, 2020 •

edited

Loading

rouault commented Mar 30, 2020

jmckenna commented Mar 30, 2020 •

edited

Loading

tveinot commented Mar 30, 2020 •

edited

Loading

jmckenna commented Mar 30, 2020

rouault commented Mar 30, 2020

tveinot commented Mar 30, 2020

gioman commented Mar 30, 2020

CRS Issue #35459

CRS Issue #35459

Comments

tveinot commented Mar 30, 2020

gioman commented Mar 30, 2020

tveinot commented Mar 30, 2020

gioman commented Mar 30, 2020

roya0045 commented Mar 30, 2020

tveinot commented Mar 30, 2020 • edited Loading

jmckenna commented Mar 30, 2020 • edited Loading

rouault commented Mar 30, 2020

jmckenna commented Mar 30, 2020 • edited Loading

tveinot commented Mar 30, 2020 • edited Loading

jmckenna commented Mar 30, 2020

rouault commented Mar 30, 2020

tveinot commented Mar 30, 2020

gioman commented Mar 30, 2020

tveinot commented Mar 30, 2020 •

edited

Loading

jmckenna commented Mar 30, 2020 •

edited

Loading

jmckenna commented Mar 30, 2020 •

edited

Loading

tveinot commented Mar 30, 2020 •

edited

Loading