Assessing a numerical cellular braided-stream model with a physical model

• Published in: Earth surface processes and landforms Vol. 30; no. 5; pp. 519 - 540
• Main Authors: Doeschl-Wilson, Andrea B., Ashmore, Peter E.
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 01.05.2005; Wiley
• Subjects: Bgi / Prodig; braided rivers; cellular automata; computational models; Fluvial forms and processes; fluvial processes; Geomorphology; Physical geography; Concept; Numerical model; Channel geometry; Model; Stream; Braided channel; Fluvial dynamics
• ISSN: 0197-9337; 1096-9837
• DOI: 10.1002/esp.1146

A. B. Murray and C. Paola (1994, Nature, vol. 371, pp. 54–57; 1997, Earth Surface Processes and Landforms, vol. 22, pp. 1001–1025) proposed a cellular model for braided river dynamics as an exploratory device for investigating the conditions necessary for the occurrence of braiding. The model reproduces a number of the general morphological and dynamic features of braided rivers in a simplified form. Here we test the representation of braided channel morphodynamics in the Murray–Paola model against the known characteristics (mainly from a sequence of high resolution digital elevation models) of a physical model of a braided stream. The overall aim is to further the goals of the exploratory modelling approach by first investigating the capabilities and limitations of the existing model and then by proposing modifications and alternative approaches to modelling of the essential features of braiding. The model confirms the general inferences of Murray and Paola (1997) about model performance. However, the modelled evolution shows little resemblance to the real evolution of the small‐scale laboratory river, although this depends to some extent on the coarseness of the grid used in the model relative to the scale of the topography. The model does not reproduce the bar‐scale topography and dynamics even when the grid scale and amplitude of topography are adapted to be equivalent to the original Murray–Paola results. Strong dependence of the modelled processes on local bed slopes and the tendency for the model to adopt its own intrinsic scale, rather than adapt to the scale of the pre‐existing topography, appear to be the main causes of the differences between numerical model results and the physical model morphology and dynamics. The model performance can be improved by modification of the model equations to more closely represent the water surface but as an exploratory approach hierarchical modelling promises greater success in overcoming the identified shortcomings. Copyright © 2005 John Wiley & Sons, Ltd.