[R-sig-Geo] Parametric/Non parametric Probability Kriging: get misclassification of risks as hazardous (false positive) and safe (false negative) in gstat?

[Zia Ahmed]

Dear All,

I did OK and Gaussian simulations of 1000 realizations of  my groundwater arsenic data after  Box-cox transformation. From OK, I created a risk map  of probability of exceeding 0.200 ppm of water arsenic by parametric approach. I need help after this analysis to solving following issues:
1)    Is it possible in gstat to get misclassification  of risks as hazardous (false positive) and safe (false negative) as describe by Goovaerts (1997)  from this parametric or non-parametric approach.
II)  After simulation of 1000 realizations of ok, Is it possible to extract prediction variance of each simulations?

I have limited knowledge  in  geostatistic as well as R. 

Thanks 

Zia Ahmed