Modeling of Coastal Morphological Processes 28.2 Principles of Process-Based Morphodynamic Modeling 617
Part C | 28.2
structures, which also interact with the profile on the
same scale as the longshore bars.
Coastal Area Models
Coastal area models are applied where a separation between longshore and cross-shore scales is not possible,
for instance in the vicinity of a tidal inlet, with a complex bathymetry with channels and shoals at varying
angles with respect to the undisturbed coast orientation.
The typical composition of an area model is as follows (see also Fig. 28.1):
A 2-D wave propagation model; by far the most
used are spectral wave models, e.g., SWAN [28.18,
19] or Mike21SW; such models solve for wave
shoaling and refraction, wind growth, whitecapping, and wave breaking and provide spatially varying field of wave properties and radiation stresses
used in the flow and sediment transport models.
A 2DH, Q3D or 3-D flow model, usually based on
the shallow water equations, solving the nonstationary flow fields and water level variations due to tide,
wind, waves, and density gradients.
A sediment transport model, usually combining bed
load transport with suspended load transport based
on the 2DH or 3-D advection–diffusion equation
or on an energetics approach, with specifically for
coastal areas important effects of wave stirring included.
A sediment balance and bottom composition bookkeeping system, which updates the bathymetry and
composition of the bottom (in case of multiple sediment fractions).
Area models are applicable at a range of scales,
from small-scale coastal engineering problems to
macro-scale evolution of tidal basins.
At the smallest scales, we are talking about the
details of the waves and wave-driven flow patterns
and resulting morphology changes around small structures such as groynes and detached breakwaters, or
the erosion of dunes where the coast is not alongshore uniform. The small grid cell sizes required to
resolve the main processes and associated small time
steps make such problems computationally intensive,
especially when predictions over longer time than some
storms are required. An important choice is whether
to treat the problem as two-dimensional, where the
focus is on the effect of the horizontal circulation patterns, or as (quasi-) three-dimensional, where vertical
circulations (undertow) are accounted for as well as
additional wave-related effects providing an onshore
component [28.20, 21]. In the first case, beach profiles
may evolve that are far from an equilibrium shape; in
the second case, there will be a tendency toward restoring a certain profile shape, but much care must be given
to tuning the cross-shore processes in such a way that
reasonable profiles develop or are maintained. At these
scales, sandbars may also be resolved, but as we discussed under profile models, obtaining the right bar
behavior is a tight balancing act. An especially difficult challenge is to model the natural evolution of
complex three-dimensional beach topography, where it
is a key to get the transitions between different beach
states right [28.22–24].
At a larger scale, for instance, in studies of harbor
extension, inlet stabilization or large-scale nourishments or land reclamations, details of processes within
the surfzone usually cannot be resolved, but the surfzone must be resolved at least enough to generate
a decent longshore current and transport; this means
a minimum of 510 grid cells across the surfzone. As
these studies are often aimed at predicting the evolution over months to years, it is important that some
cross-shore transport processes are accounted for, so
cross-shore profiles stay in a reasonable shape even
though bars cannot be represented and must perhaps
even be prevented from occurring. Part of the evolution of coastal profiles depends on the interaction with
the dunes, in terms of dune erosion and regeneration by
wind-blown transport. Such processes are typically illrepresented and some heuristic approaches to exchange
sand between the wet model domain and the dry beach
and dunes are called for.
In larger scale, longer term situations, for example
the long-term evolution of tidal basins, simulations over
time spans of centuries are now possible with the help
of new acceleration techniques, allowing us to study the
equilibrium behavior of tidal basins and the potential
impacts of the sea-level rise on such basins [28.25–
29]. Surprisingly, realistic results are obtained using
relatively simple physics, though a general tendency in
such simulations is to generate too deep and narrow
channels. Recent research, however, shows that much
better channel profiles are obtained when taking into account the spatial variation in grain sizes and especially
the tendency of sediment to coarsen with increasing
mean shear stress. When the adjacent coasts and ebb
deltas are to be properly represented wave effects must
also be included. This poses severe restrictions on the
nearshore grid resolution and drastically increases the
computational cost, typically, by at least an order of
magnitude. The applications and scales discussed in the
previous paragraphs are summarized in Table 28.1.
28.2.6 Upscaling Techniques
Morphodynamic model components each have their
typical timescales and numerical time steps. For flow
Part C | 28.2
structures, which also interact with the profile on the
same scale as the longshore bars.
Coastal Area Models
Coastal area models are applied where a separation between longshore and cross-shore scales is not possible,
for instance in the vicinity of a tidal inlet, with a complex bathymetry with channels and shoals at varying
angles with respect to the undisturbed coast orientation.
The typical composition of an area model is as follows (see also Fig. 28.1):
A 2-D wave propagation model; by far the most
used are spectral wave models, e.g., SWAN [28.18,
19] or Mike21SW; such models solve for wave
shoaling and refraction, wind growth, whitecapping, and wave breaking and provide spatially varying field of wave properties and radiation stresses
used in the flow and sediment transport models.
A 2DH, Q3D or 3-D flow model, usually based on
the shallow water equations, solving the nonstationary flow fields and water level variations due to tide,
wind, waves, and density gradients.
A sediment transport model, usually combining bed
load transport with suspended load transport based
on the 2DH or 3-D advection–diffusion equation
or on an energetics approach, with specifically for
coastal areas important effects of wave stirring included.
A sediment balance and bottom composition bookkeeping system, which updates the bathymetry and
composition of the bottom (in case of multiple sediment fractions).
Area models are applicable at a range of scales,
from small-scale coastal engineering problems to
macro-scale evolution of tidal basins.
At the smallest scales, we are talking about the
details of the waves and wave-driven flow patterns
and resulting morphology changes around small structures such as groynes and detached breakwaters, or
the erosion of dunes where the coast is not alongshore uniform. The small grid cell sizes required to
resolve the main processes and associated small time
steps make such problems computationally intensive,
especially when predictions over longer time than some
storms are required. An important choice is whether
to treat the problem as two-dimensional, where the
focus is on the effect of the horizontal circulation patterns, or as (quasi-) three-dimensional, where vertical
circulations (undertow) are accounted for as well as
additional wave-related effects providing an onshore
component [28.20, 21]. In the first case, beach profiles
may evolve that are far from an equilibrium shape; in
the second case, there will be a tendency toward restoring a certain profile shape, but much care must be given
to tuning the cross-shore processes in such a way that
reasonable profiles develop or are maintained. At these
scales, sandbars may also be resolved, but as we discussed under profile models, obtaining the right bar
behavior is a tight balancing act. An especially difficult challenge is to model the natural evolution of
complex three-dimensional beach topography, where it
is a key to get the transitions between different beach
states right [28.22–24].
At a larger scale, for instance, in studies of harbor
extension, inlet stabilization or large-scale nourishments or land reclamations, details of processes within
the surfzone usually cannot be resolved, but the surfzone must be resolved at least enough to generate
a decent longshore current and transport; this means
a minimum of 510 grid cells across the surfzone. As
these studies are often aimed at predicting the evolution over months to years, it is important that some
cross-shore transport processes are accounted for, so
cross-shore profiles stay in a reasonable shape even
though bars cannot be represented and must perhaps
even be prevented from occurring. Part of the evolution of coastal profiles depends on the interaction with
the dunes, in terms of dune erosion and regeneration by
wind-blown transport. Such processes are typically illrepresented and some heuristic approaches to exchange
sand between the wet model domain and the dry beach
and dunes are called for.
In larger scale, longer term situations, for example
the long-term evolution of tidal basins, simulations over
time spans of centuries are now possible with the help
of new acceleration techniques, allowing us to study the
equilibrium behavior of tidal basins and the potential
impacts of the sea-level rise on such basins [28.25–
29]. Surprisingly, realistic results are obtained using
relatively simple physics, though a general tendency in
such simulations is to generate too deep and narrow
channels. Recent research, however, shows that much
better channel profiles are obtained when taking into account the spatial variation in grain sizes and especially
the tendency of sediment to coarsen with increasing
mean shear stress. When the adjacent coasts and ebb
deltas are to be properly represented wave effects must
also be included. This poses severe restrictions on the
nearshore grid resolution and drastically increases the
computational cost, typically, by at least an order of
magnitude. The applications and scales discussed in the
previous paragraphs are summarized in Table 28.1.
28.2.6 Upscaling Techniques
Morphodynamic model components each have their
typical timescales and numerical time steps. For flow