[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: GSTAT: Anisotropy
Hi Simon,
> Dear all
>
> I have a data set of Cs-134 deposition in west Cumbria and I have
> been attempting to fit a variogram model to the natural log
> transformed data in order to undertake some kriging. When assuming
> isotropy I don't get a very good variogram model - I can fit a
> spherical model, but it doesn't look much better than if I assumed a
> nugget variance. Then I fitted a variogram model to the data assuming
> a direction of 225 degrees (corresponding the dominant wind direction
> perhaps) with a horizontal tolerance of 45 degrees I got a much better
> variogram model.
>
> Therefore, I would like to know whether gstat uses this direction and
> horizontal tolerance when it undertakes the kriging and estimation of
> predicted and variance surfaces. I put in the parameters associated
> with anistropy (set alpha=225.0 and set tol_hor=45;) in the command
> file and I get the same predicted and variance surfaces as when I
> krige without these parameters in the command file.
>
We had the same problem last week, I guess set alpha and set tol_hor
don't work. Use the options (p and s) in the variogram command instead,
see
http://www.geog.uu.nl/gstat/manual/node20.html
somewhere at the bottom of that page,
cheers,
Derk Jan
>
> Many thanks for any help people can provide
>
> Kind regards
>
> simon
>
> S i m o n W r i g h t
>
> Radioecology Section
> Institute of Terrestrial Ecology
> Merlewood Research Station
> Windermere Road
> Grange-over-Sands
> Cumbria. LA11 6JU
> United Kingdom
>
> Telephone: +44 (0)15395 32264
> Fax: +44 (0)15395 35941
> E-mail: s.wright@ite.ac.uk
--
Derek Karssenberg
PCRaster
Faculty of Geographical Sciences
Utrecht University
PO Box 80.115
3508 TC Utrecht
the Netherlands
Phone: (+31) 30 2532768
Fax: (+31) 30 2531145
E-mail: d.karssenberg@geog.uu.nl
http://www.geog.uu.nl/pcraster