Hi
I'm not sure to understand what you mean when you say kriging on non
planar surface.
If you are saying that you need to interpolate on to an irregular
grid, both GSLIB and GSTAT (and so R) permit
you to interpolate whatever set of points.
Bye
Sebastiano T.
At 03.14 03/04/2007, Olumide
Hi Sebastiano,
I think Olumide's question was, what to do if the underlying manifold is
not euclidean nor spheric or something regular of that kind. How would you
define the distance beween pairs of points on the surface of a potatoe, or
how would you lay a regular grid on a potatoe, which covers
Olumide,
There has been some work on geostatistics on the sphere -- the
challenges faced here included, afaics (i) to find and use distance
functions, such as great circle distance, and a suitable earth model
(sphere or rather ellispsoid) and (ii) to find and use covariance
function that are
Hi Peter
Well!!! if the point is this the things become quite difficult
If the problem is to take in to account the roughness of topography
in order to compute distances I think that with GIS technologies is
possible to find an answer.
Surely things become more complex if you have obstacles
Edzer J. Pebesma [EMAIL PROTECTED] writes:
There has been some work on geostatistics on the sphere -- the
challenges faced here included, afaics (i) to find and use distance
functions, such as great circle distance, and a suitable earth model
(sphere or rather ellispsoid) and (ii) to find and