Take the 2-minute tour ×
Geographic Information Systems Stack Exchange is a question and answer site for cartographers, geographers and GIS professionals. It's 100% free, no registration required.

I want to perform interpolation using cost distances rather than euclidean distances similar to Greenberg et al. (2011). In that paper they use r.cost in GRASS to compute cost distances for each prediction location. Unfortunately I am more familiar with ArcGIS than GRASS. Can I use the diffusion/kernel interpolation with cost surfaces in ArcGIS rather than GRASS/r.cost? Would I get equivalent results?

Greenberg, J.A., C. Rueda, E.L. Hestir, M.J. Santos and S.L. Ustin. 2011. Least cost distance analysis for spatial interpolation. Computers & Geosciences 37:272-276. http://dx.doi.org/10.1016%2Fj.cageo.2010.05.012

share|improve this question
    
I'm not familiar with that paper and its algorithm, but many years ago I used Spatial Analyst's CostDistance to perform non-Euclidean interpolations (and non-Euclidean kernel density calculations). So you definitely can do the work in ArcGIS. Whether you get "equivalent" results in the sense of exactly the same values given the same inputs will depend on minor implementation details, because there are multiple valid ways to estimate costs of movement across a grid. –  whuber Jan 25 '13 at 17:04
    
With no Answers offered in over a year, and some useful advice having already been provided as a Comment, would you be able to either write up an Answer, or edit your Question to revise it in line with your subsequent learnings, please? –  PolyGeo Apr 4 at 23:28
add comment

2 Answers 2

I asked the original question but I lost my login credentials. Let me give an update.

I worked out how to do these non-Euclidean interpolations in ArcGIS but they took a long time to process (many hours). I just recently came across an R package (ipdw) that seems to produce equivalent output 4x faster.

share|improve this answer
add comment

As @whuber answered in a Comment, he:

used Spatial Analyst's CostDistance to perform non-Euclidean interpolations (and non-Euclidean kernel density calculations). So you definitely can do the work in ArcGIS. Whether you get "equivalent" results in the sense of exactly the same values given the same inputs will depend on minor implementation details, because there are multiple valid ways to estimate costs of movement across a grid.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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