Modeling Concepts - Coastal Lagoons: Ecosystem Processes and Modeling for Sustainable Use and Development - page 287

6.3.5.4 Possible Simplification According to the Task

To Be Solved

The task of estimating the water balance for a lagoon, especially at seasonal or inter-
annual time scales, may be solved using the 0D level. The accuracy of this solution
will increase as the lagoon becomes more leaky.
For tasks involving lagoon water quality, 2D hydrodynamic and advection-
dispersion models are usually enough to resolve the seasonal variations of simulated
parameters. In that case, the model should be calibrated using some conservative
tracer time series such as the seasonal salinity evolution. A 2D time-dependent
hydrodynamic approach is also sufficient to simulate wind surges, tidal level, or
current variations in a lagoon. The tidal water exchange between a lagoon and its
adjacent coastal waters is also well simulated using a horizontally 2D approach. The
same is true for nontidal lagoons where wind surge is the main external force.
However, when a two-layer circulation is suspected in the lagoon or at its leaky
opening, the use of a 3D model is recommended. In some cases, local winds and
bathymetry combine to produce a 3D current structure in the deeper basins of the
lagoon. 40 This structure can only be simulated using a 3D hydrodynamic model ( see
Chapter 9.2, the Grande-Entrée Lagoon case study). Tasks involving short-term
forecasts following accidental oil spills definitely require a 3D approach for the
hydrodynamic and transport problems because of the possible different directions
advective transports can have at different depths. These must be simulated precisely
in order to predict the fate of the spill and optimize timely recovery operations.
Finally, tasks involving the study of water exchange between the subbasins of a
lagoon should be solved in 3D for precise short-term simulations and in 2D for
simulations involving seasonal variations.

6.3.5.5 Computer, Data, and Human Resources

Recent developments in computer hardware have shown that processing speed and
random access memories are increasing at a significant rate. Therefore, when a model
is planned to be used for a long period of time, a modeler should not restrict the
choice of models based on the model complexity (e.g., choosing 2D instead of 3D).
Instead, the choice of the model complexity should depend on the data available to
fit the model. Normally, a monitoring program should supply the data needed to
calibrate and run the model regularly. The choice of model complexity should also
depend on the qualification or the expertise of the model user. Recent models are
very complicated tools and despite the easy-to-use graphical interface presented, the
user must be able to interpret the modeling results. This requires some hydrodynamic
expertise to interpret the results in the frame of the basic physical approach that was
used. Therefore, computer hardware resources are not the major limitation for model
usage. Although qualified personnel could be a temporal problem, the main diffi-
culties lie in data availability, interpretation, and use within the model. This is a
crucial point for model selection.
Models are developing in a way that considers as many significant processes as
possible and formulates them as clearly as possible. Undoubtedly, the choice to use