[R-sig-Geo] Anisotropy in Kriging with R

Bianca Federici bianca.federici at unige.it
Wed Oct 4 15:45:04 CEST 2006


>Dear all,
>
>I'm a researcher at the University of Genova, 
>Italy, and I'm interested in interpolating 
>methods to reproduce braided rivers, hence river with multiple channels.
>We wanted to interpolate data distributed along 
>cross-sections of a river, hence with a strong anisotropy.
>
>Kriging methods implemented in R (gstat package) 
>take into account an anisotropy direction of the 
>spatial correlation between data;
>however, it's seems to us that such parameter 
>has an important role in the determination of 
>the variogram, but doesn’t influence at all the 
>interpolation algorithm that works in an isotropic way.
>
>Have you practice about this? If yes, do you conferm our conclusions?
>
>Thank you for your time
>
>         Bianca
>
>---------------------------------------------------------------------------------
>Bianca Federici
>Assegnista di ricerca
>
>DIMSET - Dipartimento di Macchine, Sistemi Energetici e Trasporti
>Facoltà di Ingegneria
>Università degli Studi di Genova
>Via Montallegro, 1 - 16145 Genova
>
>tel. 010-353 2421
>fax: 010-353 2555
>homepage: http://www.diam.unige.it/obelix




More information about the R-sig-Geo mailing list