Comparative Roles Of Roughness Reduction And Bed Erosion On Downstream Peak Discharge Increase

Wei Li,Zhengbing Wang,Huib De Vriend,Wu Baosheng
2015-01-01
Abstract:Highly silt laden flow in the Yellow River (China) tends to exhibit abnormal features during floods, among which the phenomenon of a downstream increasing discharge peak is observed in the Lower Yellow River. This has become an issue of significant interest to hydraulic engineers and scientists, also in relation to the so - called man - made floods. The increasing discharge entails an increasing load on the flood defence system and an increasing flood risk as the flood peak moves downstream. Yet, the basic mechanisms underlying this phenomenon remain to be unraveled, in order to be able to mitigate its potential damage. The many existing explanations and the complexity of this phenomenon call for a comprehensive understanding. Following the perspective of turbulence and effective bed roughness reduction in hyperconcentrated flow, we have conducted a numerical model study for the hyperconcentrated flood of August 2004 in the Xiaolangdi-Jiahetan reach. In that model, we used a simple power law relation between Manning' s n and the volumetric sediment concentration. In order to focus on the essentials of the phenomenon, the river was schematised to a 1 - D straight channel with a rectangular cross - section of constant width and with the bed consisting of uniform sediment. The feedback between the bed deformation and the turbid flow, however, is fully accounted for, in the constituting equations as well as in the numerical solutions. The model successfully reproduced the phenomenon of downstream flood peak increase when considering the hyperconcentration - induced bed roughness reduction. As the hyperconcentration lags shortly behind the flood peak, later parts of the flood wave experience less friction and can overtake the wave front. The model results also suggest that the contribution of bed deformation is much less important to the magnitude of the flood peak and its downstream evolution, at least for 'normal' hyperconcentrated floods without extreme bed erosion and sediment concentrations (e.g. 800 similar to 900 kg/m(3)).
What problem does this paper attempt to address?