This question already has an answer here:

I want to interpolate temperature for a surface defined by three coordinates: longitude, latitude and altitude. I know one option is to use universal kriging on two dimensions and use altitude as a regressor. ¿is there a way to perform ordinary kriging on three dimensions with the gstat package in R? ¿what option (3d ordinary kriging or 2d universal kriging) is likely to produce more accurate interpolation results?

marked as duplicate by Andre Silva, John Powell, Mapperz Sep 19 '16 at 16:48

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • 1
    You don't want to do this interpolation with 3D kriging: not only is it inappropriate for such data (because the surface is confined to a two dimensional manifold), it would be impossible to identify a variogram model (because you do not have suitable point-pairs aligned in near-vertical directions). – whuber Jul 15 '13 at 17:24
  • So you recommend the universal-kriging option with altitude as a regressor? – emDiaz Jul 16 '13 at 20:27
  • 1
    It would be negligent of me to recommend anything on the basis of so little information. It is clear, though, that this is not a 3D problem. – whuber Jul 16 '13 at 20:32

Browse other questions tagged or ask your own question.