Monitoring Program Design - Coastal Lagoons: Ecosystem Processes and Modeling for Sustainable Use and Development - page 321

7.4.4 M ODEL -A CCOMPANIED C URRENT D ATA S UPPLY

This special kind of regular data collecting is necessary to enable the running of
the model at any time. For this purpose, in the case where the model is already
calibrated, a limited measurement program is needed to supply the model with
current data about initial and boundary conditions only ( see Chapter 3 and Chapter
6 for details).
For example, the variations in levels at the open boundaries, discharges and
concentrations of chemicals in rivers and other external and internal sources, wind,
and initial distributions of modeled chemicals all offer possibilities to begin the model
simulation.
This accompanied (attendant) current data supply is compulsory for support of
operational modeling and simulation of transport phenomena in the case of an
emergency. For all specific purposes, the specification on data collecting during
accompanied monitoring is carefully designed according to the objectives of mod-
eling and model requirements.

7.4.5 P RACTICAL R ECOMMENDATIONS ON THE D ESIGN

OF S HORT -T ERM D ATA C OLLECTION

The problem of data collection and sampling/measurements for model calibration,
validation, and running is the key question in model implementation and in future
operational model usage. Experience shows that standard monitoring programs
executed by environmental authorities are often fragmented, both in time-space and
in the parameters set; therefore, they are very often impractical because they do not
satisfy the strict requirements on temporal and spatial compatibility for necessary
specific calibration data ( see Chapter 9, the Vistula Lagoon case study).
A special program of supplementary data collection aimed at model tuning (or
current data supply) is the first step of model implementation. A typical general
example of the specification for this type of monitoring is presented in Table 7.1.
More detailed recommendations are not given here, because specification guidelines
are to be based strictly on the specific model structure and must be prepared by an
expert in modeling in each specific case. It should be emphasized that the program
of data collecting dedicated to modeling is one of the most expensive tasks in the
process of model implementation. Some recommendations are therefore useful to
optimize it.
First, there is no way to avoid or to minimize the information about forcing
factors and initial and boundary conditions (see Chapter 3 and Chapter 6 for details).
The data must be collected in any case. The amount of data needed directly depends
on the model used (zero-, one-, two-, or three-dimensional model).
Special field experiments that aim to define the internal modeling parameters
(for example, for hydrodynamic models these are the surface and bottom drag
coefficients, or turbulent diffusion or dispersion coefficients) are rather useful, but
they are not as important as precise determination of lagoon bathymetry and geom-
etry as well as geographical positions of sinks and sources, which have first priority
together with initial and boundary conditions.