idw stands for "inverse distance weighting" which means that for
a given location to be interpolated, N known points "near" to
this point are selected and the z values for the N points are
averaged to form a z value for this location. The "average" is
weighted - points farther away are given less weight than points
nearby. The value for N is chosen by the user. The points
selected are the N points nearest the location to be
interpolated. The distance measure is Euclidean for UTM and
other planimetric grids, geodesic distance for lat-lon.
r.surf.idw and s.surf.idw differ only in that the known points
are posted to the center of gird cells in r.surf.idw whereas in
s.surf.idw the know points are not so constrained.
|
|
| I would like some information on the method used to
|generate the surface with s.surf.idw. What method was used? Is
|it similar to the "moving average" technique (Geo.
|Stats.--J.C.Davis)? Does it use the Pythagorean equation? Are
|s.surf.idw methods similar to the methods used in r.surf.idw?
|What would be a good reference for s.surf.idw?
|
|
|Regards,
|
|
|Bruce Wylie
|
|
|
Michael
While on this subject: what is a good source of kriging programs for
use with GRASS? I would like to do cokriging on data that I have in
grass site maps to produce a raster map.
Cheers,
Bob Harrington
Hydrology and Water Resources
University of Arizona