0

I have a conceptual question. I have lat-long points which I have used to extract environmental variables from. My coordinates are in WGS1984. For most of the variables, I converted the environmental data into WGS1984 before extracting the data into a matrix/vector (Lat, Long + env variables).

For a few of the variables though, I increased efficiency by converting both my lat-long points and the environmental data into a planar coordinate system. Once I extracted the value for such environmental covariate, I added the values back to this data matrix.

I ran a regression on this data and now that I'm displaying the end result, I'm unsure whether it would be proper to label my data as WGS1984. I was planning on creating a raster out of this XYZ matrix and I have to list a projection in R. Would I have to retransform the extracted data in the planar projected data back in WGS1984 before I can add it into the data matrix? If I understand correctly, these are the environmental values that would correspond to data for the lat-long in either WGS or planar?

Improve this question
4
  • If you add the data back in before unprojecting, you'll have a corrupted dataset.
    – Vince
    Commented Feb 18, 2020 at 12:41
  • Could you explain what you mean by a corrupted dataset? The data that I add back are simply just numerical values, with reasonable values in a range that I'd expect.
    – Tammy
    Commented Feb 18, 2020 at 14:44
  • If the data needs projecting, it has a spatial component; if you merge data with different spatial references in one table, you destroy the ability to distinguish between objects' locations.
    – Vince
    Commented Feb 18, 2020 at 15:01
  • Thanks @Vince. To clarify then, this is all working in R: I've transformed pt1 (WGS1984) --> pt1a (UTM) and extracting i.e. temperature in (UTM). I should now transform the dataframe of coordinates and temperature back into WGS1984 and add it to the original data matrix?
    – Tammy
    Commented Feb 18, 2020 at 15:10

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.