2

I am trying to use the Resample tool in Spatial Analyst to go from a 1kmx1km grid to a 500mx500m grid for increased resolution of my land use data. My cell size is 0.008, and I want a 0.004 cell size. I input this into the Resample tool, and it runs smoothly, but when I check the cell size after the resample, it is still 0.008.

Does anyone know why this is? I am using ArcMap 10.3.

  • How do you check your cell size? By measuring or by checking in the file metadata? If it is done by measuring, then try checking the metadata. – Mikkel Lydholm Rasmussen Nov 11 '15 at 10:30
  • If you look at the image, the resampled map looks the same, because the neighboured pixels look the same. – Iris Nov 11 '15 at 10:32
  • Did you check layer properties > source > raster information > cell size? – Iris Nov 11 '15 at 10:34
  • Hei! I have looked at the metadata in the attribute table to see if my pixels increased 4x after the resample, but it hasn't. In the metadata (looking under properties), my cell size in the resampled piece changed (I went from 0.008 to 0.004), but my pixel count is still the same as in the original. I'm using the "nearest" resampling technique and not touching anything in the "environments." Any ideas? – Alya Bolowich Nov 12 '15 at 10:02
2

I work with multiple raster layers and always try different parameters to get the best results. You can try to run the tool under "Majority" method instead of "nearest. Also, use your environments to set this:

"Processing extent" = your original raster "Snap Raster" = your original raster "Raster Analysis" = specify your Cell size

2

When working with continuous rasters, such as rainfall or elevation, try the "Cubic" Resampling Technique. This calculates the value of each pixel by fitting a smooth curve based on the surrounding 16 pixels. It produces the smoothest image but can create values outside of the range found in the source data - and in this way is best suited to continuous rasters as opposed to discrete rasters such as those based on classed attributes, eg. land cover, where surrounding pixel values may be totally independent from each other.

This method was successful for me when reducing the cell size of a rainfall raster from 2.5km to 1km so it would align with a 1km polygon grid and therefore process my zonal statistics (as table) completely. Without doing this, only 40% of the zones were processed.

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.