I have an elevation-raster in .hgt format with CRS EPSG:4326. My aim is to calculate aspect and slope which - as far as I understand - are difficult to obtain when using geographic CRS. Projected CRS are much better suited as map units are not degrees but meters (or the like).

So I tried and transformed the layer to EPSG:5677. However, the transformation produces very unpleasant artifacts (see pictures). The angles which were produced by the transformation have the potential to distort further analyses to a significant extent, I fear.

Can anyone suggest a procedure to minimise these unwished effects?

before transformation:

enter image description here

after transformation:

enter image description here

close before transformation: enter image description here

close after transformation: enter image description here

  • Did you reproject your original raster? In this answer gis.stackexchange.com/questions/143139/… you can see how to reproject a raster properly – Gerardo Jimenez Mar 10 '17 at 14:45
  • yes I did reproject it. I also tried using different resampling methods - near works worst, the others better but still far from satisfying, I think. I wonder whether there are other resampling methods not included in the standard tool? – yenats Mar 10 '17 at 15:16

Apparently, there is no "perfect" transformation (which also is obvios). However, the warp function has different resampling methods to it. In this case, "cubic" seemed to work best. The result is not exact the same as the input layer, however, the undesired artifacts can be minimized and further analysis should only be affected to a minor extent.

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.