[GRASS-user] IDW (or similar) and cost surface

[Hamish]

Giovanni wrote:
> Does anyone know of implementations of an IDW (or other
> distance weighted interpolators) which can use a cost surface
> as a supplementary weighting factor?I'm exploring solutions
> to do spatially constrained interpolation, and modelling
> barriers through cost surfaces could be a way to do it...


yes, for a low number of starting points, v.surf.icw from addons:
  http://grass.osgeo.org/wiki/GRASS_AddOns#v.surf.icw

"v.surf.icw is an IDW interpolation method using true distance cost instead of euclidean shortest distance, i.e. as the fish swims around an island not as the bird flies. This will cleanly travel around hard barriers and a cost surface map may be used to model expensive-cross barriers. Input data points do not need direct line of sight to be considered, but should be kept to less than one hundred as the module becomes very computationally expensive. A number of radial basis function options are available."

screenshot:
  http://grass.osgeo.org/wiki/File:Inlets_03_SurfSal_icw_big.png


I have not tried running it in trunk yet but the r.cost there
should be much faster.


Hamish