EP012-0013
Efficient and Accurate Long-Term One-Dimensional Morphodynamic Simulations in Alluvial Rivers Using Simplified Models - From Theory to Praxis

Tuesday, 8 December 2020
Poster
Hermjan Barneveld, Wageningen University and Research Center, Wageningen, Netherlands, Ton Hoitink, Wageningen University, Environmental Sciences, Wageningen, Netherlands, Erik Mosselman, Delft University of Technology, Delft, Netherlands and Lieke Anna Melsen, Wageningen University, Wageningen, Netherlands
Abstract:
Morphodynamic numerical simulations can be time consuming, especially in a stochastic approach when multiple long-term simulations are required to address parameter uncertainty. Several techniques have been developed to improve the efficiency of simulations, such as numerical decoupling of water and sediment dynamics and the use of a morphological acceleration factor. The scope of this research is to assess under which circumstances neglecting terms in the governing 1-dimensional Saint-Venant equations is possible, without jeopardizing the predictive capacity in terms of sediment transport and riverbed development.

Previous hydraulic studies have focused on the impact of neglecting inertial terms in the momentum equation on the propagation and damping of flood waves. Here, one of these analyses is extended to assess the applicability of simplified flow equations for river morphodynamic simulations. In a theoretical analysis of the linearized equations for water and sediment, the celerity and damping of disturbances on the riverbed are analyzed, comparing results for the full dynamic model (no terms neglected) with results from simplified models.

The Froude number and a parameter describing unsteadiness and non-uniformity of flow appear to determine whether terms in the Saint-Venant equations can be neglected while still accurately describing the propagation and damping of flood waves. For conditions with low Froude numbers and/or longer flood wave duration, a kinematic wave or a diffusive wave approach appear to be valid for hydrodynamics only. When including bed evolution, a third parameter, the dimensionless transport parameter, is introduced. For the quasi steady approach the analysis shows that the deviation from morphodynamic results with a full dynamic model increases with sediment load.

The results of the theoretical analysis have been validated with 1-dimensional numerical morphodynamic simulations for a flash flood river in India. Additional numerical runs are envisaged to verify wider applicability.