To construct the experimental semi-variogram in the direction of alignment
, the data are grouped into distance classes. Every data pair which is
separated by a distance
is used to estimate the value
.
In practice, it is easier to consider a constant tolerance
whatever the
distance
, i.e.,
constant,
.
Otherwise,
should be smaller for
smaller distances and larger for greater distances.
This grouping of data pairs into distance classes causes a smoothing of
the experimental semi-variogram relative to the underlying
theoretical semi-variogram
, cf. Figure 4.12. If the
available data
pairs separated by a distance
are used instead of those
separated by the strict distance
, then it is not
that is being estimated
but rather a linear combination of the
; more precisely, the theoretical
mean value
The significance of the effect of this smoothing over the interval
decreases as the tolerance
becomes smaller with respect to the range
of the theoretical model
to be estimated.
In practice, the following points should be heeded.