Dear brothers in kriging
They say that present efficiency of geostatistics in metal mines is (only) about 50%.
I would say : Cause
we may neglect effect of dynamic ore formation.
For the kriging equations these two correlograms
+ 1 if h = 0
a) r (h) =
+ f (h) if h > 0
a) r (h) =
+ f (h) if h > 0
+ 1 if h = 0
b) r (h) =
- f (h) if h > 0
b) r (h) =
- f (h) if h > 0
are not the same, however information about the spatial correlation is the same.
So, when the effect of dynamic ore formation is taking into consideration,
the case a) seems to be a more artificial than b). Problem is, that in the term
of variogram or covariance the kriging equations, in fact, see only case a).
One (practicing miner) should rewrite a source code in the term of correlogram
change the sign - like in the case b) and test it.
t. suslo
Yahoo! Messenger with Voice. Make PC-to-Phone Calls to the US (and 30+ countries) for 2¢/min or less.
