2

I'm currently trying to perform an analysis on a number of points and their EEZ regions in QGIS (see my earlier question here).

Here's an impression of what I've done so far:

enter image description here

In dark red, I've created a dissolved feature of a collection of line buffers around the EEZ regions. In grey, I've dissolved the relevant EEZ regions themselves.

I'd like to compare the areas of both. This is where it's getting strange. When I inspect the area of the buffered borders, it's about 960.412 square kilometers, see:

enter image description here

But when I inspect the area of the EEZs as a whole, it amounts to merely 66.739 square kilometers!

enter image description here

Clearly, the area of the EEZs as a whole should be larger than the area of the buffered EEZ borders (by a few orders of magnitude).

Questions: what's going wrong? And how can I obtain the correct areas of the buffered borders and the EEZ regions as a whole, respectively?

Max Muller
  • 375
  • 2
  • 9
  • 1
    So what CRS are your layers in? Looks like at least one of your layers is in a projected CRS (like EPSG:4326) with units in degrees: so calculating area does not make any sense. – Babel Sep 26 '21 at 14:00
  • @Babel they're both in EPSG 3857 - WGS 84 / Pseudo-Mercator – Max Muller Sep 26 '21 at 14:27
  • 3
    OK, this is a projected CRS, but lengths (and thus also areas) are heavily distorted towards the poles. Never use this CRS for any lengt/area calculations. Reproject to an equal area CRS like EPSG:6933, see: https://gis.stackexchange.com/a/412522/88814 – Babel Sep 26 '21 at 14:30
  • @Babel I've reprojected both the layers and the project to EPS: 6933, but the areas seem to remain the same. Also get a warning message saying "network request to services.arcgisonline.com timed out, any data received is likely incomplete." – Max Muller Sep 26 '21 at 14:42
  • OK, so you don't have the data locally? Than you can't reproject, I guess... Can you save your data locally (to a file like Geopackage)? – Babel Sep 26 '21 at 14:54
  • @Babel the warning message has disappeared. I guess I can save the data in a Geopackage. Could I perhaps upload it to some website, so you can inspect it? The areas haven't changed after the reprojections, as far as I can tell – Max Muller Sep 26 '21 at 15:05
  • You can't upload it here, but to some filesharing service and then post the link here – Babel Sep 26 '21 at 15:12
  • @Babel sent an invitation to your UZH address for a google drive map with the relevant files – Max Muller Sep 26 '21 at 15:35
  • When I reproject your two layers to EPSG:6933 and calculate the area in sq km with $area/1000000, I get the following values: 2’561’668.9 for 20 km buffer and 9’625’592.2 for eezs. This seems more realistic, doesn't it? Are you sure you correctly reprojected your layers? See: https://gis.stackexchange.com/a/383437/88814 and https://gis.stackexchange.com/a/392388/88814 – Babel Sep 26 '21 at 16:12
  • @Babel it does seem more realistic, I'll try the $area/1000000 formula. Though I still suspect there should be a bigger difference in the areas, now they only differ be a factor less than four – Max Muller Sep 26 '21 at 16:17
  • @Babel And yes, I indeed did the reprojection wrong, made the classic mistake of setting the CRS to something different – Max Muller Sep 26 '21 at 16:20
  • The difference is strange, indeed. However, your buffer layer contains a lot of geometry errors. And are you aware that your eez layer contains polygons outside of Europe as well (French Guyana, around Madagascar): you really want to keep them? – Babel Sep 26 '21 at 16:52
  • Also consider: you did the buffers in a CRS that does not represent distances well - so probably the buffers need to be re-created. As in your first question, I don't really understand why you buffer the coastline and not the whole polygons and than subtract the landmass with a difference tool. – Babel Sep 26 '21 at 16:54
  • @Babel good points, I'll look into this more carefully in a few days. Thank you for all your help! – Max Muller Sep 26 '21 at 19:02

0 Answers0