[R-sig-Geo] nested variograms?

[Edzer Pebesma]

Dave Depew wrote:
> A quick question for experienced gstat users....
> Can nested variograms be fit using gstat? if so, is it simply adding 
> to an existing variogram structure?
>
> Thanks
>
 > library(gstat)
Loading required package: sp
 > data(meuse)
 > coordinates(meuse) = c("x", "y")
 > v = variogram(log(zinc) ~ 1, meuse)
 > fit.variogram(v, vgm(1, "Exp", 300))
  model     psill    range
1   Exp 0.7186595 449.7664
 > fit.variogram(v, vgm(1, "Exp", 300, 1))
  model     psill    range
1   Nug 0.0000000   0.0000
2   Exp 0.7186528 449.7583
 > fit.variogram(v, vgm(1, "Exp", 300, add.to=vgm(1, "Nug", 0)))
  model     psill    range
1   Nug 0.0000000   0.0000
2   Exp 0.7186528 449.7583
 > fit.variogram(v, vgm(1, "Exp", 300, 1, add.to=vgm(1, "Exp", 10)))
Warning: singular model in variogram fit
  model psill range
1   Exp     1    10
2   Nug     1     0
3   Exp     1   300
 > fit.variogram(v, vgm(1, "Exp", 300, 1, add.to=vgm(1, "Exp", 10, 
add.to=vgm(1, "Exp", 50))))
Warning: singular model in variogram fit
  model psill range
1   Exp     1    50
2   Exp     1    10
3   Nug     1     0
4   Exp     1   300

Of course the fitting fails badly for this poor example, but I guess you 
get the idea. I think there is no limit to the number of nested 
structures, but consider that feature as pretty useless.

-- 
Edzer Pebesma
Institute for Geoinformatics (ifgi), University of Münster,
Weseler Straße 253, 48151 Münster, Germany.  Phone: +49 251
8333081, Fax: +49 251 8339763  http://ifgi.uni-muenster.de/