[R-sig-Geo] gstat: Cokriging with unusual neighbourhood choice - is this possible?

[Edzer Pebesma]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Dear Ralph,

I guess you'd need to define three variables,

concA, precA and precB. Then, you can let precA and precB have
different neighbourhood specs, but make them statistically "identical"
by specifying the same variograms for precA and precB as well as for
its cross variogram, and use the merge option to have them share a
single mean value. (see ?gstat ; there's an example of this in
demo(pcb) ).

On 07/23/2013 11:21 AM, Ralph Mettier wrote:
> I'm looking at two variables, concentration and precipitation,
> measured at two networks of stations:
> 
> network A measures both variables at few (~30), widely spaced
> stations network B measures only precipitation at many (~1800)
> densely spaced stations
> 
> a decent correlation between the two variables is expected, making 
> cokriging an attractive choice. Ordinary cokriging in gstat works,
> using only concentration from network A and precipitation only from
> network B. However, for quality assurance, I am trying to reproduce
> previous results obtained using a different, commercial,
> geostatistics tool.
> 
> Currently, I am using nmax (and maxdist) to control the points in
> the neighbourhoods (16 concentration and 32 precipitation), but
> what I would ideally need to do is the following:
> 
> For concentration, the neighbourhood should contain the 16 nearest 
> measurements from network A, as is the case now. For precipitation,
> the neighbourhood should contain the 16 nearest measurements from
> network B and the 16 nearest measurements from network A. This
> doesn't seem possible using the 'nmax' option in gstat.
> 
> I have tried splitting concentration and precipitation from network
> A into two seperate gstat objects and then using cokriging with
> three variables, but the results are unsatisfying and I'm fairly
> certain this is not the way to go.
> 
> Any hints or pushes in the right direction would be greatly
> appreciated.
> 
> Best Regards Ralph
> 
> [[alternative HTML version deleted]]
> 
> _______________________________________________ R-sig-Geo mailing
> list R-sig-Geo at r-project.org 
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
> 

- -- 
Edzer Pebesma
Institute for Geoinformatics (ifgi), University of Münster
Weseler Straße 253, 48151 Münster, Germany. Phone: +49 251
83 33081, Fax: +49 251 8339763 http://ifgi.uni-muenster.de
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBAgAGBQJR7lzRAAoJEM1OCHCtOnfx3pUH/27TvcVqFg50DPXkqKjX7c9N
9RE36Sy1lFd/WG8ky6VK3Wb+M3kNfNh1ZsVyK/zjYwjD7QswGgsSl7ZGgt29GrEJ
S0271WagmRYFLG7iHD4F00NOLffe2d7wl9Go5t/paYh1chjI8JQFWwSUJy9J9u3y
0yBwZ3SWhqULaijdbVlSVsdSv09PPuirz/j1KIQqmzWD+E/Rtno64oPUuBVPWV2K
WYTNvstYQ6R5pdMrOj5yPKKSPkzqwj5W7ohRYgIsIXnZGCV8uQUnF5UmLPOwln6G
+rjrS3rSGlBYqmcgXPpvCOfs9fexiuZL+VL21TDg/std3lWI5ob6akCYDhkKpaw=
=ceyk
-----END PGP SIGNATURE-----