[R] kriging problem(?)

[robert.ptacnik at niva.no]

Hei,
I have two spatial datasets Sa and Sb, both with lat-lon coordinates and
from same geographic area, but from different localities within the area
(independent samples). Sa is biotoc data, Sb is some environmental
parameter (fertility). I 'know' that Sb affects Sa, but wonder on which
scale. I tried different interpolations by creating different grids of Sb
(e.g. 20x20 and 100x100 points using 'interp' ...) and found that a more
coarse grid did a better job in predicting Sa than a high resolution grid.
However, I wonder if there aren't more elegant ways, e.g. using kriging.
Just to avoid misunderstandings: I do not just want to get a spatial grid
of Sa and Sb, but I want to find on which Scale Sb affects Sa.
cheers,
Robert,


 Robert Ptacnik
 Norwegian Institute for Water Research (NIVA)
 Gaustadalléen 21
 NO-0349 Oslo
 FON +47 982 277 81
 FAX +47 221 852 00



----------------------------------------------------------------------------------------------------------------------
NIVAs hovedkontor har flyttet til nye lokaler i CIENS - Forskningssenter
for miljø og samfunn; Gaustadalléen 21, 0349 Oslo. Meld deg på vårt
nyhetsbrev på www.niva.no).