Monitoring objectives, pollutantsources,
get dictate much of the design. The need for moni-
toring a spatial control and the need to quantify
conditions before, during, and after land treatment
Monitoring
comprise the remainder of design decisions. Fi-
Program Design
nally, the manager should verify that the monitor-
ing committee's design
address the monitoring
objective.
The monitoring program design is the framework for sampling, data analysis, and
the interpretation of results. Typically, the objective of a NPS pollution control
project is to document changes in water quality that are related to the NPS controls.
A time series
Monitoring both the water quality and the land treatment/land use in a project can
provide valuable feedback regarding the impact of land management on water
document changes in water
quality. This chapter emphasizes land treatment and water quality monitoring
quality due to land treatment.
designs to meet the objectives of detecting trends and/or direct impacts of land
treatment on water quality; in addition, objectives for evaluating current condi-
Measurements should either
tions and problem documentation are discussed.
be taken at regularly timed
A time series must be obtained to document changes in water quality due to land
intervals (e.g., every 7 or 74
treatment. Measurements should either be taken at regularly timed intervals (e.g.,
every or 14 days) or for specified periods and for a sufficient length of time using
days) or for specified periods
comparable, consistent methods.
and for a
length of
The components of a time series are both deterministic and random. The determin-
istic component changes in a predictable manner and is assumed or known without
time using comparable,
error (e.g., time, seasonal cycles, or treatment strength). The random component is
consistent methods.
measured with error and consists of unexplained factors that hinder the detection of
the trend. To detect a trend, the random component and complex deterministic
factors such as cycles (e.g., climatic or life cycle), and the dependence of one
observation on the next (serial correlation), must be taken into account.
Improvement from NPS control occurs gradually, and few, if any, agricultural
NPS control studies have shown a step trend in the receiving water. More often the
change is incremental and subtle, and visual detection of a change can result in
For trend detection,
false conclusions: claiming progress where none has occurred, or failing to detect
monitoring objective should
small but real improvement. Therefore,
monitoring to detect changes in water
quality due to land treatment requires an experimental design to isolate land
be translated into a testable
treatment effects.
statistical hypothesis.
For trend detection, the monitoring objective should be translated into a testable
statistical hypothesis to provide structure to the experimental design. The null
hypothesis states that no change is expected. The monitoring survey is designed,
using the principles of experimentation, to test the null. If the design is sound and
statistical testing shows the null hypothesis to be false, then a change can be
4.7
Previous Page
