Geoscience Reference
In-Depth Information
5.3.6.2 1-D channel meandering model
Simulating channel meandering processes using a 1-D numerical model is quite
difficult, because of the strongly three-dimensional features of flow and sediment
transport in meandering channels. However, ignoring the channel meandering pro-
cesses may induce significant errors in the simulation of long-term flow and sediment
transport in alluvial channels. Therefore, many investigators have studied this prob-
lem and proposed empirical or semi-empirical methods to account for the influence of
channel meandering on sediment transport.
From a computational point of view, one needs to know the flow path and its
potential change during the simulation using a 1-D model. One approach is to adopt
empirical relations to calculate the flow path of a meandering channel under given
flow and sediment conditions. Some channel regime theories may be used.
The other approach is to use meandering migration models, which may be kinematic
or dynamic. The kinematic migration models relate the migration rate to channel
width, curvature, etc. Examples are Ferguson (1983) and Howard and Knutson
(1984). The dynamic migration models solve the simplified dynamic equation of flow
to estimate the flow properties in the meandering channel, and relate the bank erosion
rate to the excess velocity or shear stress at the outer bank. Examples can be found in
Johannesson and Parker (1989) and Odgaard (1989).
Because of the truly three-dimensional flows and highly complex soil properties in
meandering channels, most of the current 1-D channel meandering models are only
applicable in simple cases. Further study on this problem is needed.
5.3.7 Overall procedure for 1-D decoupled flow and
sediment calculations
The individual modeling components in the fully decoupled and semi-coupled models,
which decouple flow and sediment calculations, have been introduced in previous
sections. The overall calculation procedure for the 1-D decoupled unsteady model
consists of the following steps:
(1) Calculate the unsteady flow using the Preissmann scheme and the double sweep
method based on initial channel geometry;
(2) Calculate sediment transport, bed change, and bedmaterial sorting fromupstream
to downstream, using the known flow conditions;
(3) Determine the bed change due to bed material consolidation, if needed;
(4) Correct channel geometry by allocating the bed change along the cross-section;
(5) Calculate bank erosion and mass failure, if needed, and
(6) Return to step (1) and conduct the calculations for the next time step, based on
the new channel geometry, until all time steps are finished.
The calculation procedure for the 1-D decoupled quasi-steady model is almost the
same as the above procedure, except that the standard step method is used for the
quasi-steady flow calculation in step (1) and the time interval (in hours or days) used
in the quasi-steady model is usually longer than the time step (in minutes) used in the
unsteady model.
