0

I would like to resample a number of rasters with a small cell size (i.e., 30 m) to a larger cell size (i.e., 100 m) using either ArcMap or ArcGIS Pro. My new, larger cells will each cover 3.33x3.33 of the original cells, but the resampling window for the different techniques is inappropriate and appears to be hard-coded into the geoprocessing tool.

  • Bilinear (continuous data): resamples the 4 cells nearest to the center of the new cell. This window is too small.
  • Cubic (continuous data): resamples the 16 cells nearest to the center of the new cell. This window is too large, and will result in an undesireable level of smoothing.
  • Majority (categorical data): ArcMap uses a 3x3 window and ArcGIS Pro uses a 4x4 window. One is a little too small, and one is a little too large, but they are both close enough to be OK.

Is there a way to specify a user-defined window (e.g., the 10 closest cells) for resampling continuous raster data?

3

In the ArcGIS universe, there are a couple things you can look into:

  1. If you're familiar with ArcGIS's "raster functions," there is a Resample function, which provides additional resampling options beyond what the geoprocessing tool offers, including an average option that returns "the average value of all involved pixels". This doesn't let you specify the number of neighbors to consider, but might be close enough to what you're looking for.
  2. Keep in mind that the raster resampling algorithms are really just special cases of spatial interpolation. If you want maximum control over the interpolation, you could convert your raster to points and use a different spatial interpolation method, such as IDW. The ArcGIS IDW tool allows you to specify a number of points (e.g., the 10 closest points) in place of a search radius. Requisite disclaimer: it is of course your responsibility to determine the appropriateness of IDW for interpolating your data (but that goes for any of these approaches).
  • Adding to @eweber here are some links to understand what type of interpolation is better for what. link link – Daniel May 31 at 16:13

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.