Improved method for identifying Manning’s roughness coefficients in plain looped river network area

Abstract

Manning’s roughness coefficient(n) is considered a key parameter in a one-dimensional (1D) hydrodynamic model. However, it is highly variable and time- and site- dependent. Further, identifying proper n values is not easy, especially in plain looped river network areas. Therefore, a more systematic approach is needed. This study proposes a coupled optimization-simulation model to systematically estimate the spatial distribution of n values. The particle swarm optimization (PSO) algorithm and InfoWorks Integrated Catchment Modelling (ICM) software were integrated to solve the objective function and hydraulic process, respectively. Crisscrossing rivers were partitioned into river reaches that were each assumed to have a uniform and constant Manning’s roughness coefficient according to their network topology and cross-section variation. In addition, a sensitivity analysis was implemented to determine the weights of measured data from different gauging stations, and a large difference in the spatial distribution of the sensitivity index illustrated the importance of identifying the weights of multiple stations. Then, a systematic approach was applied to estimate the n values in the Changshu Grand Polder Area (CGPA), which is crisscrossed by 150 rivers, under the water diversion stage. The calculation statistics and efficiency indicated that the proposed method performs well for model calibration.

Keywords

• manning’s roughness coefficient
• plain looped river network area
• particle swarm optimization (pso)
• sensitivity analysis
• latin hypercube sampling (lhs)
• infoworks integrated catchment modelling (icm)