Quantcast SAMPLING CONSIDERATIONS

 
  
 
will require knowledge of the range of lake depths, the number of significant inflows, the presence of
large or unique tributary embayments, and (if possible) pre-existing knowledge of spatial trends in the
lake.
In the absence of statistical design, or where specific needs are restricted to certain areas of a
lake, the number of depths of observations is important. Too many depths will require large amounts of
sampling time and too few depths may not resolve the depth structure sufficiently to address the
question. This becomes more true for deeper lakes in which significant investments of time are made
for relatively few depth <profiles'. If pre-existing knowledge allows, it is possible to vary the sample
depth increment to economize on effort and time. For example, increments of one meter in depth in the
region of the thermocline can yield to five- or ten-meter increments in deeper waters if the temperature
does not significantly change. However, this may lead to problems of data management or data
analysis if the variety of sampling is great.
In small, morphometrically simple lakes a single sampling location may be sufficient to
adequately characterize lake. In morphometrically complex lakes, many sampling locations will be
necessary to perform such characterizations and the procedures for data analysis become important as
well.
Environmental problems that are acute or which only occur during certain seasons may require
one or a few repeated sampling efforts. However, trends (especially those over long times) require
regular repeated efforts. These time-series efforts can lend themselves to specialized statistical
techniques associated with time series analysis. In those analyses the ability to resolve certain
frequencies of phenomena depend not only on repeated observations but also on the frequency with
which those observations are made. In plain terms, for example, diel processes require more than one
or perhaps two observations per day, and perhaps many more if the process in question occurs rapidly
at a particular time during the day (such as the behavior of schooling fish at dawn). There are still many
lake ecological phenomena that have not been adequately resolved for lack of adequate sampling.
There are powerful new techniques for rapid sampling for special applications. For example,
the incorporation of real-time GPS positioning with ADCP profiling or with rapid in situ methods
provides high-resolution capability for spatial sampling over both short and long distances. Likewise
satellite and aircraft-based sensor technologies can provide detailed characterizations of lake surface
characteristics when they are employed with adequate field ground-truth measurements.
Such approaches are exciting but still limited in application. Remote sensing is generally
accepted but only for lake surface characterizations. In as much as some correlation can be made to
characteristics at depth, such surface characterizations can be used to provide broad, accurate
descriptions of spatially complex lakes. However, such approaches provide only short-term (snapshot)
characterizations and longer-term work often employs remote continuous monitoring techniques.
2.3-5

 


Privacy Statement - Copyright Information. - Contact Us

Integrated Publishing, Inc.