Parameter Optimization in Flood Routing
Peter R. Wormleaton, M. Karmegam
Abstract
Peter R. Wormleaton, M. Karmegam
Abstract
The writers demonstrate how the geometric and hydraulic properties of river reaches, which are required in flood routing when using the Saint‐Venant equations, may be identified using optimization methods. It is suggested that these optimization methods may prove an attractive alternative to the costly and time‐consuming survey work or uncertainty, or both, that often accompany the estimation of numerical river model parameters. The writers adopt the four‐point finite difference scheme of Amein and Fang (2) to solve the Saint‐Venant equations. They report on an investigation into the selection of time and space increments in order to limit the finite‐difference error in the solution. The optimization process involves minimizing the errors in depth and discharge of the downstream routed hydrograph. Two objective function criteria are compared, namely minimax and least‐squares. Four optimization parameters are used, two representing channel geometry and two its hydraulic properties. Five flood events are optimized and generally the two optimization methods give consistent results, although there are differences between winter and spring floods.
OpenAlex reports 14 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The writers demonstrate how the geometric and hydraulic properties of river reaches, which are required in flood routing when using the Saint‐Venant equations, may be identified using optimization methods. It is suggested that these optimization methods may prove an attractive alternative to the costly and time‐consuming survey work or uncertainty, or both, that often accompany the estimation of numerical river model parameters. The writers adopt the four‐point finite difference scheme of Amein and Fang (2) to solve the Saint‐Venant equations. They report on an investigation into the selection of time and space increments in order to limit the finite‐difference error in the solution. The optimization process involves minimizing the errors in depth and discharge of the downstream routed hydrograph. Two objective function criteria are compared, namely minimax and least‐squares. Four optimization parameters are used, two representing channel geometry and two its hydraulic properties. Five flood events are optimized and generally the two optimization methods give consistent results, although there are differences between winter and spring floods.
Key concepts: Mathematical optimization, Hydrograph, Routing (electronic design automation), Channel (broadcasting), Optimization problem, Flood myth, Applied mathematics, Computer science