5

I've run into this issue a few times before in the past and I'd finally like to get a better idea of what's actually going on.

When reprojecting a layer, some polygons that are located around edge of the map extent will get stretched completely across the map. Is this indeed because of their location on the edge of the extent, and does this dramatically affect the accuracy of the polygon? Or is it more of a rendering artifact that does not seriously affect the underlying geometry?

The pink coral data in EPSG 4326 is the original data: epsg 4326 data

The reprojected (to equal area) orange coral data in EPSG 3410 is here: epsg 3410 data w issue

Edit: Data w Pacific central meridian It does appear to be a 180th meridian issue. epsg 4326 data centered on Pacific epsg 3410 data centered on Pacific

  • It looks as though some of your polygons might be crossing the 180th meridian (or whatever meridian is the opposite of your central one). This post explains that issue and how to solve it. – FSimardGIS Sep 26 '16 at 23:34
  • Updated the original post w pictures showing maps with Pacific centering. Do you know if this is simply an aesthetic issue or if it does in fact mess with the area of the polygons? – srha Sep 26 '16 at 23:51
  • After seeing your screen captures, i agree that it must be a 180th meridian-related issue, since your polygons display correctly when you use another central meridian. In that case, i would make sure that your polygons are clipped at the 180th meridian before you run the reprojection of your layer. – FSimardGIS Sep 26 '16 at 23:52
  • Actually, what must have happened is, when you ran the reprojection, the polygon vertices that were located to the east of the meridian got "transferred" to the western side of the new projection, thus creating this weird artefact. Indeed if you try to calculate, say, polygon areas or lengths using the new projection, these polygons might give you a wrong result. But if they have been clipped beforehand, it should be reprojected correctly. – FSimardGIS Sep 26 '16 at 23:57
  • Ok yes this makes sense. Thank you. If you post your last comment as an answer I can mark it as correct. – srha Sep 26 '16 at 23:59
6

Actually, what must have happened is, when you ran the reprojection, the polygon vertices that were located to the east of the 180th meridian got "transferred" to the western side of the new projection, thus creating this weird artefact. Indeed if you try to calculate, say, polygon areas or lengths using the new projection, these polygons might give you a wrong result. But if they have been clipped (see this related post) beforehand, it should be reprojected correctly.

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.