Geoscience Reference
In-Depth Information
determined based on bed load sediment flux and
the net result of deposition and entrainment. At
every time step the bathymetry is updated with a
morphologic acceleration factor of 60 (Lesser
et al
., 2004; Roelvink, 2006) which implies that
constant hydrodynamic conditions (including
sediment transport) are assumed for 60 subse-
quent time steps. This allows for increased com-
putational efficiency.
The stratigraphy is defined according to the
multi-layer concept of Ribberink (1987). Individual
beds are specified by thickness and sediment com-
position (sand/silt by volume). On top of the beds
a transport layer of fixed height (here, 0.2 m) allows
exchange of sediment with the water flow. The
lowest bed is the base layer, which contains the
sediment from beds with numbers that exceed
the given bed tracking capacity. The substrate
below the base layer is assumed to be unerodible.
Together, the individual beds (0.1 m) and the sub-
strate represent the stratigraphy. It is a vertical
array of cells which records the sedimentary com-
position of morphodynamic elements such as bars,
ridge and distributaries.
An extensive sensitivity analysis has been
performed to test the robustness of the simulations
and optimise the values of several parameters with
respect to formulations of sediment transport and
sediment characteristics, bed roughness and mor-
phological acceleration (Hillen
et al
., 2009). The
optimised settings are summarised in Table 1. The
geometry and stratigraphy of a delta modelled
with a comparable Delft3D setup (Geleynse
et al
.,
2010) is used as an initial condition at the start of
the simulations. Wave reworking is investigated
by subjecting this predefined delta, which formed
under mere fluvial forcing, to a time-invariant
quiet-weather wave climate.
N
Free-surface
short waves
0
-2
-6
2 km
-10
Fig. 2.
Model domain and initial morphology as input to
subsequent wave reworking. MWL is beneath Mean Water
Level.
2650 kg m
−3
, a dry bed density of 500 kg m
−3
and a
d
50
of 50 µm. The sand sediment fraction has a
specific density of 2650 kg m
−3
, a dry bed density
of 1600 kg m
−3
and a d
50
of 125 µm.
Sediment transport for both suspended and
bed-load is calculated with the sediment transport
formulation TRANSPOR2004 (van Rijn, 2007a;
van Rijn, 2007b). This transport formulation pro-
vides one unified framework for both types of
sediment transport and is applicable for a broad
range of sediment sizes (fine silt to coarse sand)
subject to varying conditions. It accounts for cohe-
sive effects of fine sediments (<62 µm), as well as
for the combined effects of waves and currents,
particle-particle interaction, bed slope effects,
flocculation and hindered settling. Current-
induced and wave-induced sediment transport
can be scaled with (calibration) transport factors.
Since wave-related transport is often over-
predicted in Delft3D (Lesser
et al
., 2004), the
wave-related transport factors in our simulations
are scaled accordingly (0.3). Bed shear stresses are
calculated with van Rijn's roughness predictor
(van Rijn, 2007a), which is based on four types of
roughness; grain size roughness, wave-related bed
form roughness, current-related bed form rough-
ness and the apparent roughness. The apparent
roughness is the bed roughness resulting from
wave-current interaction. Bed level changes are
SCENARIOS
The setup of the numerical experiments represents
an abandoned delta lobe following an upstream
avulsion. Sea-level is presumed to be stable. For
the sake of simplicity the simulated initial delta is
a fully fluvial-dominated delta, implying no wave
influence during its formation. However, with the
abrupt decrease of river discharge, the delta enters
the retrogradational phase of the delta cycle
(Roberts, 1997) and waves become the dominant
process. The simulations start at that exact
moment. Such a model setting is comparable to
Search WWH ::
Custom Search