[R-sig-Geo] Why correlation coefficient between observed value and predicted values very low after cross validation[gstat]?

[Uzzal]

I am trying to perform ordinary Kriging using gstat package on a hourly particulate matter(PM) concentration dataset. my dataset contains PM concentration for 1 hour of 107 sites. You can download dataset from here. After performing ordinary kriging, I did cross validation. but I got the correlation coefficient between observed values and predicted values is ver low (.15~.30). I checked many combination of model (Gau,Exp,Sph), width, cutoff, intial parameter value but still correlation coefficient is low. Am I doing any mistake in my code? Is there anything to do, to get high correlation coefficient? My r code using gstat package: library(sp)
library(gstat)
library(rgdal)seoul030101.cv  #Coordinates
coordinates(seoul030101.cv)  #Variogram modeling
seoul030101.cv_var #Cross Validation                            
cv_pr #Correlation coefficient#correlation observed and predicted, ideally 1
cor(cv_pr$observed, cv_pr$var1.pred) # correlation predicted and residual, ideally 0
cor(cv_pr$var1.pred, cv_pr$residual) #other statistics# mean error, ideally 0:
(mean(cv_pr$residual))#rmse(rmse # Mean square normalized error, ideally close to 1
(mean(cv_pr$zscore^2)) Thanks in advance.Orpheus 




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20151023/f33dbd77/attachment.html>