10 Applications of the Inverse Problem of Pollution Propagation
361
probabilities, and a method for 2D optimization) into one entity. Although construction of a single modelling system is possible, we have intentionally highlighted
the interfaces and connections between the separate basic steps so that the user may
easily replace each component with a more convenient or more elaborated one. The
essence of the method does not depend on the properties of a particular model for
each single step. The presented approaches highlight the potential of such a multistep modelling system for the preventive planning of maritime activities. They also
reveal several instructive features of the outcome of the technique and particularly
the importance of a feasible choice of the target function.
It is important to emphasize that implementations of the technology described in
this chapter are based on several simplifications, and the validity of some of them
is not obvious. They mostly account for the impact of surface-current-driven propagation of pollution on the optimum solution. Realistic applications have to account
for not only the impact of wave- and wind-induced transport of contaminants but
also complicated physical and chemical processes affecting their fate (ASCE 1996).
These questions will be partially addressed in Chap. 11.
The constructed example fairways are only conditionally optimized. The presented idealized examples (in which only the current-driven propagation is accounted for) signify, however, that the impact of current-driven transport may add
a significant contribution to the marine-induced hazards in certain coastal sections.
A major simplification that might not be entirely justified is neglecting the potential impact of subgrid-scale processes in applications using the RCO model and the
DMI/BSHcmod model. Doing so may considerably alter the statistics of trajectories,
especially for longer time intervals.
The presented material suggests that eddy-permitting models with a grid step
exceeding half the local baroclinic Rossby radius are suitable for a quick check of
whether or not any potential gain from the presented method is feasible. However,
higher-resolution simulations using eddy-resolving models are definitely necessary
for detailed planning.
Although the instantaneous values of the basin-wide probability of coastal hits
and particle age reveal substantial spatio-temporal variations in different sea areas
and/or in different years and seasons (Fig. 10.3), their cumulative values converge
quite rapidly to a fixed level for a wide range of resolutions of the ocean model.
This convergence suggests that these asymptotic values characterize certain intrinsic
combinations of the geometry of the basin and internal current structure, and perhaps even could be interpreted as new measures of the vulnerability of the nearshore
of water bodies to offshore traffic accidents for the chosen length of time window.
The (spatial) standard deviation of the probability and particle age indicates the
magnitude of their variations across the water body and thus serves as a natural indicator implicitly characterizing the potential benefit from the smart positioning of
dangerous activities for the particular sea area.
The very limited dependence of several discussed measures and properties on
the resolution of the underlying hydrodynamic models needs some attention. The
probable reason is the multi-step averaging procedure of short-term transport features (over time intervals exceeding the typical turnover time of mesoscale eddies
361
probabilities, and a method for 2D optimization) into one entity. Although construction of a single modelling system is possible, we have intentionally highlighted
the interfaces and connections between the separate basic steps so that the user may
easily replace each component with a more convenient or more elaborated one. The
essence of the method does not depend on the properties of a particular model for
each single step. The presented approaches highlight the potential of such a multistep modelling system for the preventive planning of maritime activities. They also
reveal several instructive features of the outcome of the technique and particularly
the importance of a feasible choice of the target function.
It is important to emphasize that implementations of the technology described in
this chapter are based on several simplifications, and the validity of some of them
is not obvious. They mostly account for the impact of surface-current-driven propagation of pollution on the optimum solution. Realistic applications have to account
for not only the impact of wave- and wind-induced transport of contaminants but
also complicated physical and chemical processes affecting their fate (ASCE 1996).
These questions will be partially addressed in Chap. 11.
The constructed example fairways are only conditionally optimized. The presented idealized examples (in which only the current-driven propagation is accounted for) signify, however, that the impact of current-driven transport may add
a significant contribution to the marine-induced hazards in certain coastal sections.
A major simplification that might not be entirely justified is neglecting the potential impact of subgrid-scale processes in applications using the RCO model and the
DMI/BSHcmod model. Doing so may considerably alter the statistics of trajectories,
especially for longer time intervals.
The presented material suggests that eddy-permitting models with a grid step
exceeding half the local baroclinic Rossby radius are suitable for a quick check of
whether or not any potential gain from the presented method is feasible. However,
higher-resolution simulations using eddy-resolving models are definitely necessary
for detailed planning.
Although the instantaneous values of the basin-wide probability of coastal hits
and particle age reveal substantial spatio-temporal variations in different sea areas
and/or in different years and seasons (Fig. 10.3), their cumulative values converge
quite rapidly to a fixed level for a wide range of resolutions of the ocean model.
This convergence suggests that these asymptotic values characterize certain intrinsic
combinations of the geometry of the basin and internal current structure, and perhaps even could be interpreted as new measures of the vulnerability of the nearshore
of water bodies to offshore traffic accidents for the chosen length of time window.
The (spatial) standard deviation of the probability and particle age indicates the
magnitude of their variations across the water body and thus serves as a natural indicator implicitly characterizing the potential benefit from the smart positioning of
dangerous activities for the particular sea area.
The very limited dependence of several discussed measures and properties on
the resolution of the underlying hydrodynamic models needs some attention. The
probable reason is the multi-step averaging procedure of short-term transport features (over time intervals exceeding the typical turnover time of mesoscale eddies