Toward Marine Environmental Predictions in the Mediterranean Sea Coastal Areas
349
considers also the loss of predictability due to open boundary condition effects.
This factor can be as large as the internal nonlinearity of the flow field in limiting
the predictability of the flow field (Walstad and Robinson, 1990).
The first assessment ofthe predictability ofthe flow field was done without considering the effects of atmospheric forcing on the oceanic mesoscale flow field. For
the coastal areas, the uncertainties in the definition of the atmospheric forcing, with
respect to those in the definition of the initial condition, could be the most severe
limit to the predictability of the flow field. In addition to that, the memory mechanism of the coastal circulation is most likely coming from the lateral boundary conditions especially ifboundaries are put on the shelf slope or even in deeper parts of
the basin. The imprecise knowledge of these lateral boundary conditions could be
another serious limitation to the predictability time scales ofthe flow field. Furthermore, ne sting techniques could induce adjustment processes in the interior flow
field that could severely limit the predictability of the shelf areas. Lastly, if the
coastal area is dominated by river discharges, the specification of the characteristics of the river runoff can be seen as a serious limitation to the predictability time
of the flow field.
In conclusion, the coastal environmental prediction problem has a multiplicity of
system components that could limit the predictability time. They are: 1) the limited
predictability ofthe atmospheric forcing directly influencing the coastal dynamics;
2) the lateral boundary condition uncertainties, meaning with that both the open
boundary conditions and river runoffuncertainties, affecting the long term memory
of the system and the major driving mechanisms; 3) the initial adjustment processes to the nesting techniques; 4) the initial condition specification for all the
dynamical variables of interest; 5) the flow field nonlinearities.
In order to reduce some or all the possible sources of unpredictability for the
coastal flow field, the coastal environmental problem has to consider the design of
system components for both the large and coastal scales. In Fig. 17.5 we illustrate
the two major system components, e.g., observing and modeling, for the large and
shelf scales, with data assimilation being part of modeling.
Starting from the observing system components, the large-scale part has been
developed in the past twenty years and it consists of interlinked networks of different measurements providing high quality observations of the surface and ocean
interior dynamical variables. As said in the introduction, our attention is focused on
the permanent and sustained observing systems and thus Fig. 17.5 lists only these
kind of measuring networks. The major global systems are satellite ba sed with sensors indirectly measuring sea surface temperature, sea surface height, surface
winds, and surface color. For ice covered areas these satellites retrieve also ice
extent and thickness. Today, the satellite data sets provide the highest quality synoptic information at global scale and they are the essential component of the future
permanent observing system for the oceans. However, these global satellite systems do not provide high quality data in near coastal areas due to problems of signal interference with the nearby land areas. This is a technological problem that
should be solved in the near future.
349
considers also the loss of predictability due to open boundary condition effects.
This factor can be as large as the internal nonlinearity of the flow field in limiting
the predictability of the flow field (Walstad and Robinson, 1990).
The first assessment ofthe predictability ofthe flow field was done without considering the effects of atmospheric forcing on the oceanic mesoscale flow field. For
the coastal areas, the uncertainties in the definition of the atmospheric forcing, with
respect to those in the definition of the initial condition, could be the most severe
limit to the predictability of the flow field. In addition to that, the memory mechanism of the coastal circulation is most likely coming from the lateral boundary conditions especially ifboundaries are put on the shelf slope or even in deeper parts of
the basin. The imprecise knowledge of these lateral boundary conditions could be
another serious limitation to the predictability time scales ofthe flow field. Furthermore, ne sting techniques could induce adjustment processes in the interior flow
field that could severely limit the predictability of the shelf areas. Lastly, if the
coastal area is dominated by river discharges, the specification of the characteristics of the river runoff can be seen as a serious limitation to the predictability time
of the flow field.
In conclusion, the coastal environmental prediction problem has a multiplicity of
system components that could limit the predictability time. They are: 1) the limited
predictability ofthe atmospheric forcing directly influencing the coastal dynamics;
2) the lateral boundary condition uncertainties, meaning with that both the open
boundary conditions and river runoffuncertainties, affecting the long term memory
of the system and the major driving mechanisms; 3) the initial adjustment processes to the nesting techniques; 4) the initial condition specification for all the
dynamical variables of interest; 5) the flow field nonlinearities.
In order to reduce some or all the possible sources of unpredictability for the
coastal flow field, the coastal environmental problem has to consider the design of
system components for both the large and coastal scales. In Fig. 17.5 we illustrate
the two major system components, e.g., observing and modeling, for the large and
shelf scales, with data assimilation being part of modeling.
Starting from the observing system components, the large-scale part has been
developed in the past twenty years and it consists of interlinked networks of different measurements providing high quality observations of the surface and ocean
interior dynamical variables. As said in the introduction, our attention is focused on
the permanent and sustained observing systems and thus Fig. 17.5 lists only these
kind of measuring networks. The major global systems are satellite ba sed with sensors indirectly measuring sea surface temperature, sea surface height, surface
winds, and surface color. For ice covered areas these satellites retrieve also ice
extent and thickness. Today, the satellite data sets provide the highest quality synoptic information at global scale and they are the essential component of the future
permanent observing system for the oceans. However, these global satellite systems do not provide high quality data in near coastal areas due to problems of signal interference with the nearby land areas. This is a technological problem that
should be solved in the near future.