3

I am working on a land use/land cover map of Lebanon with QGIS. This shapefile has a .prj file, but apparently QGIS fails in recognizing the projection, as it results shifted northeastwards.

See below (purple, the correct location of Lebanon; sky blue, the LULC map as positioned by QGIS according to the .prj file)

enter image description here

Here's the .prj content

PROJCS["stereolev",GEOGCS["GCS_levant",DATUM["D_levant",SPHEROID["Clarke_1880_IGN",6378249.2,293.46602]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]],PROJECTION["Double_Stereographic"],PARAMETER["False_Easting",0.0],PARAMETER["False_Northing",0.0],PARAMETER["Central_Meridian",39.15],PARAMETER["Scale_Factor",0.999534104],PARAMETER["Latitude_Of_Origin",34.2],UNIT["Meter",1.0]]

Also, the .prj is not associated to any EPSG already listed in the QGIS database.

According to epsg.io, in Lebanon three different EPSG are used

22700 Deir ez Zor / Levant Zone

22770 Deir ez Zor / Syria Lambert

22780 Deir ez Zor / Levant Stereographic

Changing the CRS to 22780 did not result in any change, while 22700 or 22770 apparently seemed to be the appropriate CRS. "Apparently", because some kind of bias emerged at closer inspection, some kind of clockwise rotation.

enter image description here

enter image description here

enter image description here

Is there a way to fix the issue (in QGIS or, even better, with R)? Any idea on what is all about?

1

Have you tried to reproject your layer? I can't make an attempt without an example but might be worth a try. Read in your original shapefile and use the .prj string as you did at the beginnings which resulted in the extreme NE shift. Go to Processing tab>toolbox>Reproject Layer Try a few different EPSGs and see if it projects the layer where you expected it.

Another option is try some of the other EPSGs in QGIS that feature the Clarke 1880 (IGN) ellispoid such as EPSG 4011.

edit: I converted your central meridian and latitude of origin to DDMM and the closest I came to figuring out what projections your proj code is defining is something based on the Roussilhe Oblique Stereographic projection. There is a reference to this in this document (go to top of page 3). I know its not an answer but it might point you in the right direction.

  • Thank you James, I already tried to reproject the layer, EPSG 22770 did the trick... almost, since it overlays the layers in the right place, except for the clockwise rotation. I tried with EPSG 4011 but it wasn't correct. I'll go in-depth with the Roussilhe Oblique Stereographic projection, thank you! – Quechua Jul 5 '18 at 9:20
  • Good to hear it but is this still not the same problem you started with? If you can post a link to where you got the original shapefile (if possible) I can take another look. – James Keating Jul 5 '18 at 9:47
  • Reprojection is not the right tool if the data is misplaced. In fact, reprojection does not change the location, only the coordinates (assuming on-the-fly-reprojection is enabled). What you relly need to do is assign a different CRS to the data, because the current CRS is definitely wrong. – AndreJ Sep 5 '18 at 10:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.