|
ABSTRACT
Bedload transport
is an essential component of river dynamics and estimation of
bedload transport rate is important for practical computations of
river morphological variations because the transport of sediment
through river channels has major effects on public safety, water
resources management and environmental sustainability. Numerous
well-known bedload equations are derived from limited flume
experiments or field conditions. These time-consuming equations,
based on the relationship between the reliability and
representativeness of the data utilized in defining variables and
constants, require complex parameters to estimate bedload transport.
Thus, a new simple equation based on a balance between simplicity
and accuracy is necessary for using in small rivers. In this study
the easily accessible data including flow discharge, water depth,
slope, and surface grain diameter d50 from the three
small rivers in Malaysia used to predict bedload transport. Genetic
programming (GP) and artificial neural network (ANN) models that are
particularly useful in data interpretation without any restriction
to an extensive database are presented as complementary tools for
modelling bed load transport in small streams. The ability of GP and
ANN as precipitation predictive tools showed to be acceptable. The
developed models demonstrate higher performance with an overall
accuracy of 97% for ANN and 93% for GP compared with other
traditional methods and empirical equations.
A three-dimensional
numerical model was applied to study the bed morphology and bedload
transport of the junction of Ara and Kurau rivers for short term event
and for high flow with 100 ARI. SSIIM2 a 3D, k-epsilon
turbulence computational fluid dynamics model with an adaptive,
non-orthogonal and unstructured grid has been used for modelling the
hydrodynamic of confluence. The numerical model was tested against field
data from Ara-Kurau confluence. Satisfactory agreement was found between
computed and measured bedload and bed elevation in the field. The study
indicates that numerical models became a useful tool for predicting the
bedload transport rate in such complex dynamic environment. The results
have demonstrated that the short term hydrologic variability can
considerably influence the morphodynamics of Ara-Kurau channel
confluence and for the different flow conditions the bedload transported
near to edge of shear layer. The coincidence of the shear layer that was
generated the considerable turbulence indicated that the increasing
turbulence levels contribute substantially to the required increase in
bedload transport capacity. The simulation results showed the grain size
distribution on the bar at the downstream junction corner is remarkably
constant and the particle size in the upstream part of the bar is more
affected by the changes in flow conditions than the downstream end where
the median diameters not varied during the period.
|