Remote continuous monitoring, in turn, depends on adequate maintenance and calibration of the
sensor arrays. Such arrays can be limited to single devices or they can extend to large numbers of
recording devices. The amounts of data produced by all of these technologically advanced approaches
are often enormous and can tax data management hardware, software, and personnel time.
18.104.22.168 Temporal Dependence
The changes occur in lakes through the year demand different observations. For example,
seasonal stratification demands intensive observation at depths where density changes most rapidly.
This is usually in the region of the thermocline. In shallow, well-mixed lakes sampling would not need to
be as intense. Exceptional situations, however, exist where thermal patterns and chemical patterns do
not coincide. These situations are more common than not and parameters such as dissolved oxygen
concentrations or specific conductance are useful indicators of such chemical trends. In some
situations, pH is also useful. Adequate sampling in these situations requires either prior knowledge of
expected chemical trends, application of field judgement based on measured in situ trends, or repeated
sampling after laboratory analysis has identified depth ranges in which chemical trends are important. In
the best studies all of these work together to formulate the most rational sampling effort.
Where inflows have periodic or sporadic changes, loading estimates may require intensive
sampling of short-term events. Two common situations exist: upstream dams have releases dependent
on water needs and power demands, and streams flowing in from the watershed experience base flows
punctuated by storm high flow events. These two situations are quite different and require different
strategies for sampling. Dam inflows are somewhat predictable and at most times the magnitudes of
flow are known. Sampling these is a simple matter of coordinating sampling effort with expected
Streams pose different problems. Where active gaging stations exist, records of flow are
available and high flows can be anticipated on the basis of meteorological information. Sampling is
problematic and involves either labor-intensive manual sampling or expensive mechanized sampling.
Either approach can be expensive and can risk missing the events of interest.
During base flow conditions, sampling can be much less demanding. Samples merely need to
be collected at dependable intervals and coordinated with the flow data. Upstream dams usually
contribute their inflows almost solely during operation. Non-operational releases are not desired and
measures are usually taken to minimize them. Therefore sampling is best during operation and flow data
is usually immediately available.
Where flow data cannot be provided through other sources, inflowing streams must be rated
and gaged on a temporary basis. This is a significant effort and requires mapping the stream cross-
section according to shape and current velocity under different stage conditions. An equation can
thereby be derived to predict flows at various stages. Once established, a simple staff gage can be