Open swalterfub opened 3 years ago
Not sure if this should be a feature request to support non-earth crs rather than a true bug.
It's a bug -- we definitely have support for non earth crs, we just have some UI gremlins when using them...
Still valid on QGIS 3.22.3.
Confirm, having the same issue in QGIS 3.22 (Linux Mint 20.3, GDAL 3.4)
Confirmed, same in QGIS 3.22.3 / proj 8.1.1 on OSX
Confirmed, having the same issue in QGIS 3.16.8 (Windows 10)
this seems to be a more severe problem than only "non-earh-CRS".
On MacOS, i get this constantly. Fresh project, loading standard OSM XYZ layer, adding some DEM data that gets correctly positioned. Once I try to merge the DEM tiles, i get the "no transform available" error. No matter if LTR or most recent version. On Windows, the same procedure works flawlessly.
Describe the bug When using non-Earth projected data, I get "no transform available" error messages. After setting up a suitable custom geographic CRS, two error messages remain (they are probably just warnings, but marked as CRITICAL). This problem has been posted on QGIS-users and development suggestions are provided there. Please follow the thread https://lists.osgeo.org/pipermail/qgis-user/2021-March/048406.html as well.
How to Reproduce Load in some non-Earth data, e.g. from Mars in Sinusoidal projection: https://maps.planet.fu-berlin.de/mex4/h1104_0000.ihs.53.tif The project CRS switches from "EPSG:4326" to "unknown CRS", in fact to the "custom CRS" of the data. Two errors are thrown:
CRITICAL No transform available between SINUSOIDAL MARS and EPSG:4326
andCRITICAL No transform available between SINUSOIDAL MARS and Custom CRS
Following Even Rouault's advice, I manually create a suitable geographic CRS:GEOGCRS["MARS 3396 geographic", DATUM["D_MARS", ELLIPSOID["MARS",3396000,0, LENGTHUNIT["metre",1, ID["EPSG",9001]]]], PRIMEM["Reference meridian",0, ANGLEUNIT["degree",0.0174532925199433, ID["EPSG",9122]]], CS[ellipsoidal,2], AXIS["longitude",east, ORDER[1], ANGLEUNIT["degree",0.0174532925199433, ID["EPSG",9122]]], AXIS["latitude",north, ORDER[2], ANGLEUNIT["degree",0.0174532925199433, ID["EPSG",9122]]]]
The lat/lon conversion does then work, but the (marked as critical) warning messages remain (this time complaining about transform between the new projection and Custom CRS, and about the new projection and EPSG:4326, which are explained by Nyall Dawson in the thread mentioned above as results fromQGIS and OS versions
Additional context
https://lists.osgeo.org/pipermail/qgis-user/2021-March/048406.html