1

I have downloaded the Corine Land Cover map 2018 (raster 100m) from Copernicus Services, which covers the European Continent. As I only need to work within a certain area, I need to clip the original data from another dataset that I have, to work just with my AOI. The work that I will need to do is about spatial and statistical analysis and modelling.

My question is concerning projections when clipping the data from the original dataset, and the resulting output rasters.

The original dataset projection is ETRS 1989 LAEA (EPSG 3035). My AOI projection is ETRS 1989 Transverse Mercator. If its correct to say, I may add that the latest is the one that should be appropriate because is local. The issue comes next:

a) If keeping the original projection while clipping, the output dataset will have the pixels in a diagonal orientation, which is not what I want, and the Sum of the cell count will be 8,925,616. enter image description here

b) If transforming the original projection into the local projection that is used in the region, the output dataset will have the pixels in the vertical orientation, which is what I want, but the sum of the cell count will be higher: 8,926,281
enter image description here

What is the best approach for the following scenario and why?

1

You should make sure that you specify exactly what you want, meaning:

  1. projection should match the project aoi's projection
  2. the bounding box should match the project aoi's bounding box
  3. the pixel resolution should match the project aoi's pixel resolution
  4. since you are using thematic data, you should use nearest neighbor resampling.

If you warp a raster from projection A to projection B, this will mean that you can have a different amount of cells for each thematic class compared to the input. In general, this is fine, since the Warp ensures that the output matches the input spatially as much as possible.

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.