0

I am trying to do a very simple interpolation with soil nitrate data. I wanted to do IDW because points close to each other should be most similar. I used the IDW interpolation function in ArcMap (image 1), and got some strange results.

The layer is stretched/elongated on the vertical axis (image 2). In fact, when I try kriging and other methods from different packages, its still elongated. I'm still new to this and thought I understood the input variables.

Is there a variable or parameter that I am missing which over emphasizes the vertical relationship?

Info on data source: Geographic Coordinate System: GCS_WGS_1984, Datum: D_WGS_1984, Prime, Meridian: Greenwich, Angular Unit: Degree

Info on tiff layer: Spatial Reference: Polar_Stereographic, Datum: D_WGS_1984

Edit: When the data are projected using polar stereographic, they no longer correlate to the map and do not represent the original sampling scheme (images 3 and 4)

Image 1: IDW inputs

Image 2: IDW with nitrate

Image 3: Data plotted with Polar Stereographic compared to data with WGS 1984

Image 4: Data with Polar Stereographic

  • Could you edit your question to include: coordinate system of your MXD and coordinate system of the source data? Depending on how IDW calculates distances - this looks like the IDW is fine but simply looks distorted due to the project map projection. – JimT Sep 16 at 8:24
  • Thank you very much for your comment. I added the coordinate system information. The spatial reference for the mxd is polar stereographic (samples from Antarctica), but I had to import the chemistry data as x-y data with WGS 1984. – MAD Sep 16 at 13:28
  • There's a fair possibility I've got this the wrong way round then. The IDW looks to be using distances in degrees for the weighting - which at these latitudes will result in elongation. Try projecting your source data to Polar_stereographic first before running the IDW. – JimT Sep 16 at 14:10
  • When I project the source data to polar stereographic, the data points are no longer on the map layer and plot clustered (incorrectly) in a completely different region. The map layers come from the Polar Geospatial Center. – MAD Sep 16 at 14:42
  • If you reproject the data and it ends up in a different place then it has been reprojected incorrectly. – JimT Sep 16 at 15:00
0

The advice from JimT was spot on. The coordinates collected with the standard GPS were in WGS 1984. Even though the map file was also in WGS 1984, the spatial reference was in Polar Stereographic. It was not enough to just switch the data over to Polar Stereographic. The lat and long data needed to be transformed to WGS84 Antarctic Polar Stereographic. I used the tool on the Polar Geospatial Center website (https://www.pgc.umn.edu/apps/convert/). Once I did this, the data look great.

WGS84 Antarctic Polar Stereographic

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.