Statistics - Field Sampling: Principles and Practices in Environmental Analysis - page 169

soil surface appears to be the same. If a sandy, textured soil is bordered by a field with
boulders and rocks just below the surface, the characteristics of that area will not be the
same as the field to which kriging is being applied. Application to this rocky area will
give inaccurate results.
The above situations should not present a problem if sufficient samples are taken and
used in the kriging estimation. It is essential that there be a number of component data
points representing the entire area of interest [12-15].
I know of one example in which such a situation resulted in a catastrophic dam
collapse. A dam was built on an underlying rock layer. Unfortunately there was an area
of unconsolidated rock at a distance less than the range that was used in sampling, and
thus was not detected. This eventually led to undercutting of the dam and its collapse.
This resulted in massive flooding downstream.
6.15. CONCLUSIONS
Statistics is an essential tool in sampling. It is used to determine if two components have
the same or different levels of a component of interest. It is used to determine values of
interest at unsampled positions. It is also used to estimate the number and positions of
samples needed. Geostatistics is useful in estimating contours and missing values from
data sets. Various geostatistical methods are used in commercial modeling programs and
so are valuable in the modeling of environmental situations. It is highly recommended
that a statistician be consulted in using statistics in a sampling plan. It is also highly
recommended that appropriate computer statistical programs be used in carrying out all
statistical analyses.
QUESTIONS
1. Using a hand calculator or a computer generate 10 random numbers.
2. List and explain the four types of hypotheses.
3. Calculate the median and mean for the TPH values given for areas C1 through C6 in
Figure 6.1.
4. Calculate the standard variation of the TPH values given in sampling areas A1
through A6 in Figure 6.1.
5. Using the t statistic determine if the areas C1 and A6 represent the same or different
populations.
6. Determine if the value 346 ppm for TPH in sample area B2 is an outlier or is from
the same population for the area.