I have a dataset consisting of irregularly-spaced points (x,y,value) that represent an area of ocean that contains some land. I need to interpolate these to a regular grid, ensuring that the interpolation routine does not consider data points that are on the opposite site of a land barrier (e.g. the other side of a peninsular). I understand that this means that I need to insert "breaklines" in the data, and I have polygons of the coastline that I can use for this.

Do any R packages allow for spatial interpolation using breaklines? Ideally I think that natural neighbour interpolation would be the most appropriate for my dataset, but I suspect that the breaklines are more important than the interpolation method.

  • When posting this question I wasn't sure whether it would be best here or on Stackoverflow. Please feel free to migrate if the community feels that it would be better over there. Thanks!
    – Flyto
    Commented Jul 17, 2014 at 14:46

1 Answer 1


The ipdw package accounts for barriers during interpolation. It is an application of Inverse Distance Weighting using path distances. You can find the vignette here.

You might also consider non-Euclidean kriging using the geoRcb package on github. I have not tried it but it looks promising.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.