[R-sig-Geo] Spatial Analysis Doubts

[Nuno]

Hi you all,
 
I am studying the spatial overlap between two species. I want to:
 
 - Determine the spatial distribution of each of the two species using a
geostatistical approach by estimating the variogram and then apply the
kriging and/or cokriging.
 
 - Use the estimates predicted for each species to determine the ratio, to
evaluate de spatial distribution of the ratio estimates and then obtain the
overlap degree.
 
My questions are:
 
1 – What are the procedures that you recommend to test if the process is
stationary? I just assumed that the process is stationary?
 
 
2 - In my work there is clear east-west anisotropy. My questions regarding
this are:
 
i) Is there any test to evaluate its significance other than graphically? 
 
ii) Being a strong anisotropy it is more advisable to study the spatial
distribution, of the variable under study, in that direction?
 
3- Are there any cases where the Simple Kriging is more advisable than the
other forms of Kriging?
 
4- Besides Cross-validation residual and Cross validation z-scores, are
there any test to evaluate the adequacy of different models? Is there any
coefficient like AIC criterion to evaluate and choose the best-fitted model
between several models or the decision is mainly done by graphical analysis? 
 
5- Which is the package in R more fit to my study? geoR or gstat?
 

Thanks a lot for your help,
Best regards,
Nuno Veiga



--
View this message in context: http://r-sig-geo.2731867.n2.nabble.com/Spatial-Analysis-Doubts-tp7581802.html
Sent from the R-sig-geo mailing list archive at Nabble.com.