For spatial data, the semivariogramgenerally rises with the increase
of distance between neighboring samples that are correlated and then
stabilizes at the sill, indicating the distance (a) beyond which samples
are considered to be spatially independent. Variance that exists at a
scale shorter than the field sampling is found at zero lag distance and
is known as the nugget effect (C
). The nugget effect reflects the uncertainty
caused by sampling errors or short-scale variability.
0
The fitted model provides information about the spatial structure
and the input parameters for kriging. Therefore, kriging guarantees a
minimum variance unbiased prediction as well as an estimate of the
prediction variance (Fu et al., 2010). In this study, Ordinary Kriging is
used to map the spatial distribution of heavy metals (Goovaerts, 1999;
Liu et al., 2008).