0

I am trying to apply IDW interpolation in a situation where I have multiple clusters of points, and all points in each cluster have identical coordinates. My understanding is that different IDW implementations have different ways to deal with repeated measurements (see for instance this and this posts)

Currently I am using the R package gstat v2.0-6, and I can't really figure out how it addresses this issue.

So my question is: what is a proper/accepted way to deal with points with identical coordinates when using IDW? I guess that my options are:

  • average all values from repeated observations before running the IDW algorithm
  • add some minute spatial noise so that the distance between any two points is always > 0
  • ?

Switching to other interpolation methods is probably not at option, given the very limited extent of the data set.

Improve this question
6
  • This depends. What are your measurements of? If they truly are measurements of the same point, how far apart are they? Does sampling the first time influence results of the second sampling? Perhaps you could add some more context about the points.
    – geoscience123
    Commented Feb 4, 2021 at 14:38
  • Thank you for your comment. The data set is composed of zoological measurements from different individuals. All samples from a given sampling site have identical coordinates although realistically they had various degrees of spatial spread across the landscape. More precise location data, as well as any chronological information, is not available unfortunately. This is the reason why I'm looking for a generic yet acceptable approach.
    – Zanon
    Commented Feb 14, 2021 at 17:22
  • But what are you trying to get out of interpolation? It seems like the wrong method for only 2 points.
    – geoscience123
    Commented Feb 14, 2021 at 20:15
  • Sorry, that was an excessive simplification just to have an easily manageable example. I have multiple clusters of points distributed over an area of several km2. I have amended my original question. The interpolation serves primarily an aesthetic purpose. It is meant to fill the map with smooth color transitions instead of using just colored points. Regardless of this superficial usage, I would still prefer to learn how to apply it correctly.
    – Zanon
    Commented Feb 15, 2021 at 22:06
  • Gotcha. I'm not sure what is accepted in your field, but if there is no significant time between your samples, I would say just find a basic average and use that.
    – geoscience123
    Commented Feb 15, 2021 at 22:10

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.