Geoscience Reference
In-Depth Information
10
Sampling Designs for
Environmen
tal Monitoring
Trent McDonald
10.1 Introduction
The statistical part of environmental monitoring designs contains three
components: the spatial design, the temporal design, and the site design. The
spatial design dictates where sample sites are located in the study area, the
temporal design dictates when sample sites are visited, and the site design
dictates what is measured at a particular site and how. These design compo-
nents are largely, but not completely, independent. For example, methods in
site design can substantially influence sample size and the population defi-
nition which in turn influence the spatial and temporal designs. However,
the site design may simply require selection of a point as an anchor for field
measurements, and the spatial design can select this point without knowl-
edge of the specifics of the field measurements. Likewise, the temporal
design can dictate when points are revisited without knowing their location
or the specifics of field measurements. It is therefore useful and relatively
easy to study general-purpose spatial designs without knowledge of the
temporal or site designs. This is the approach taken by this chapter.
When designing the spatial component of an environmental monitoring
study, the overriding purpose should be to select a sample of geographic
locations that will eventually allow analyses that satisfy the study's objec-
tives. In most cases, these objectives include making valid scientific infer-
ences to all parts of the study area. In other words, the results of the final
analysis (e.g., point estimates, confidence intervals, distributional statements,
etc.) should be accurate and apply to the intended population.
145
