4

I am trying to view the World EEZ polygon from Marine Regions.org, but centered on the Pacific Ocean.

To centre the EEZ map on the Pacific ocean I followed these steps. The custom CRS from this is: +proj=robin +lon_0=-150 +x_0=0 +y_0=0 +ellps=WGS84 +datum=WGS84 +units=m +no_defs

So far, so good.

However, I would then like to dissolve the shapefile so that the parts of EEZ's which were originally split by the projection separated at 180, but which actually share attribute information, are one part, without the split where they would be separated under the original projection. This can be seen in the EEZ of New Zealand and nearby EEZs.

I have tried:
1) Eliminate sliver polygons
2) Dissolve, by a common attribute (EEZ Name, e.g. 'New Zealand').

I also tried save the shapefile to the new projection, but when I try to use the dissolve or sliver processes on this, the temporary file QGIS creates a warning saying 'projection not identified, reprojected to WGS84', and the polygons do not appear.

This answer does not work for me, which may be because I am using a custom projection which QGIS does not recognise.

enter image description here

  • The download page also offers a 0-360 degree shapefile, it does not have split lines at 180°. – AndreJ May 30 '17 at 8:23
  • True! But I am also curious for other cases where the 0-360 is not available. – Amroco May 30 '17 at 17:59
  • @Amroco I can't test this solution, so I post it as a comment: could you try to use the v.clean GRASS tool from the Processing Toolbox with the rmdangle parameter? – mgri May 31 '17 at 14:00
2

Similar to my post in the linked topic, the data needs some handwork before dissolving works:

  1. Check for geometry errors
  2. Split polygons that cross the antimeridian of your projection (30°E)
  3. Reproject the data to Robinson
  4. Perform multipart to single part
  5. Check the nodes where the 180° meridian cuts the borders. Left and right points might not match, so snap them with a tolerance of 10 pixels
  6. Dissolve the polygons

You might need to assign the Robinson projection again between the steps 3 to 5.

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.