Geoscience Reference
In-Depth Information
Fig. 14.31
Histogram and basic statistics. All 2 m Au composites (
top left
); High-grade HAZ zone (
top right
); Low-grade HAZ zone (
bottom
left
); and SAP zone (
bottom left
)
clustered class means, and the percentage of total contained
metal for each class with respect to the total quantity of metal
(QM) in the database, expressed as
(g/t)*l,
where
g/t
is the
Au grade, and
l
is the composite length. Note how the high-
est grade classes defined for all three domains contain a sig-
nificant percentage of the total metal in the zone.
The cell-declustering technique (Deustch
1989
) was
used to assess the degree of clustering in the deposit. Fig-
ure
14.33
shows the histogram and basic statistics of the
cell-declustered 2 m composites for the Haz-Hi zone, and
should be compared to Fig.
14.31
. The clustering effect is
significant, particularly for the higher grade units.
threshold, although for some of the very high thresholds,
with few composites in the class, the variogram model was
assumed to be a pure nugget effect.
Since there is a large number of indicator variogram mod-
els to consider, one for each threshold of each estimation
domain, only a few model summaries are shown here as ex-
amples. The main conclusions are:
1. As expected, variograms are more continuous for lower
indicators, with lower nugget effects.
2. Nugget effects increase and ranges decrease with increas-
ing thresholds, i.e., there is de-structuring of the vario-
gram for higher thresholds.
3. Indicator variograms at the 3.0 g/t or 5.0 g/t thresholds
show an almost pure nugget effect, and therefore all var-
iogram models for higher thresholds are assumed to be
also pure nugget effects. It should be noted that these in-
dicator thresholds are important, even though their vario-
gram models are pure nugget effect, because they provide
critical control of very high-grade values.
14.2.4
Indicator Variograms
The indicator thresholds were used to define the correspond-
ing set of indicator variograms. In most cases, a total of 37
experimental variograms were obtained for each indicator
