Statistics - Field Sampling: Principles and Practices in Environmental Analysis - page 162

Hawaii — 14,000-

32,000

Transects on island of Hawaii, 1-to-2-km
spacing, 0-15 cm deep

Yost et al.
[11]

FIGURE 6.4 Decrease in TPH concentration from a point source.

Several other possibilities can exist; however, only two will be discussed here. One
would be that the concentration or characteristic does not change over the distance of a
transect or sampling blocks. For instance, it might be possible that within experimental
error the pH of a soil does not change over the entire area being sampled. In this case the
characteristic would follow a model with a slope of 0; that is, there is no change in the
value with a change in either the x or the y direction. In an absolute or realistic sense this
would not happen. What actually happens is that the pH varies around some mean and the
means at all places are not significantly different from each other.
The second would be that the characteristic changes in a regular fashion over the
distance of the transect. In Figure 6.5 soil bulk density changes in a regular pattern
because of wheeled traffic across the field. This type of graph may result from
measurement of any environmental characteristic or quantity, whether it is chemical or
physical.
In the types above, if the expected values are dependent on position, then universal
kriging, which uses variograms (see below) is used. It might also be noted that at a finite
variance both correlograms and variograms contain the same information.