ordinary kriging

[<leuchner@gsf.de> (Michael Leuchner)]

Dear Gstat users,

We are using ordinary kriging to interpolate annual average precipitation 
data. We assumed that the obtained values for our input data points would be 
the same after kriging than before. But the results vary quite a lot at some 
points. The values generally are too low for high input values and vice 
versa. What is our mistake in order to receive the same values for our 
precipitation stations after kriging as before ? Are we using the wrong 
method ?

Here is our command file:

data(var): 'D:\krig_1.txt', x=1, y=2, v=4, min=4, radius=0.33, average;
variogram(var): 41322.5 Pen(72421.4) + 10113.3 Nug(0);
mask: 'geo_mask.map';
predictions(var): 'krig_83.map';


If anyone has got an idea we would appreciate it a lot:

Regards,

Michael Leuchner.