Morphological Changes due to River Bank Protection

When bank protection works are constructed along an eroding outer bank, it is often observed that the bed along the outer bank becomes substantially deeper, sometimes causing the failure of the bank protection works themselves. Despite our awareness of this deepening phenomenon, its causes and quantitative estimation is still a matter of investigation. Two mechanisms are said to be responsible for this channel deepening: one is reduction of the input of sediment entering from eroding banks and the other is channel narrowing due to continued pointbar growth and accretion of the opposite bank. In order to clarify the main cause and to develop an estimation method, a simple one-dimensional (axi-symmetric) mathematical model was developed, which accounts for the channel widening/narrowing. The model includes a schematized expression of the bank erosion/accretion mechanism in addition to the traditional flow and sediment transport sub-models. Since the model includes two important aspects - bank erosion and accretion -, first of all their effects on the behaviour of the model was examined. Secondly, an attempt was made to apply the model to a specific bank protection project along the Brahmaputra River in Bangladesh. Through these analyses and application, it is concluded that the model can simulate qualitatively two important aspects of the behaviour of natural rivers: o a wide, shallow and relatively steep channel forms when the banks are easily eroding, while a narrow, deep and mild-slope channel forms when the banks are resistant to erosion; o bank stabilization makes rivers deeper, and yields a milder slope, as an extreme case of banks which are resistant to erosion. This capability of the model also suggests its high potential for predicting the channel width in alluvial rivers. For slowly migrating rivers , the model is able to predict the morphological changes due to bank protection works. For rapidly migrating rivers, however, the model seriously overestimates the channel widths, and underestimates the bank migration rate and the outer bend scour depth. Although the results obtained are quite promising, much more research will be needed to arrive at a good quantitative prediction method. For this purpose, the following further research is recommended: o examination of the bank accretion mechanism and estimation of the bank erosion/accretion parameters through model calibration; o axi-symmetric analysis including 1) suspended load, 2) variation of roughness coefficient, 3) discharge variations and 4) secondary flow convection; o two-dimensional analysis (or even three-dimensional analysis for the flow field) including 1) suspended load, 2) streamwise variations of flow and sediment distribution in the beginning of bends, 3) variation of roughness coefficient, 4) discharge variations and 5) secondary flow convection; o establishment of simple prediction methods or a design diagram for the outer bend scour depth.