"Hoogerwerf, ir. M.R." wrote:
>
> The gstat manual (appendix A.3) mentions the possibility to do block kriging
> for irregularly shaped blocks, using the area command/option.
>
> It mentions that the 'file' should contain the points that describe the
> shape. I'm not sure what it expects. Is it :
>
> 1). A "mask" map that has a value for all cells within the irregular block
> and No Data for all other cells
Yes. At your option, defined as a mask map or as an ascii file with the
points
that discretize the area.
Then, there are two options:
a) no prediction locations defined
b) prediction locations defined (using mask:... or data():... )
In case a) the area coordinates point to the prediction area (`block'),
in
case b) they do not (and they will get centered around (0,0,0) anyway)
> 2). The vertices of the irregular shape specified by x,y coordinates in a
> GeoEas type data file.
No.
Good luck,
--
Edzer