Automating HEC-RAS and Linking with Particle Swarm Optimizer to Calibrate Manning’s Roughness Coefficient

摘要

Abstract Hydraulic models have a substantial role in the simulation of rivers due to their high accuracy and low cost. One of the most practical hydraulic models is HEC-RAS capable of simulating all flow conditions in watercourses. Floods occurring in rivers are highly dependent on Manning's Roughness Coefficient (MRC). Its optimization, calibration, and uncertainty analysis are necessary. To this end, HEC-RAS should be automated and linked by optimization models so that it can seek to find the optimal values ??using an iterative process. In this research, HEC-RAS was automated in MATLAB2019 and linked with Particle Swarm Optimization (PSO) and Mont Carlo Simulation (MCS). The sensitivity analysis of PSO was performed, and its optimal coefficients were determined. The MRC calibration was done in the Shahab River in Hamadan province (Iran). The results showed that the MRC in the three distinguished reaches (from upstream to downstream) were respectively obtained as 0.061, 0.057, and 0.040 in the main channel and 0.069, 0.059, and 0.046 in the floodplain. Comparing the obtained values from optimization and estimated values by traditional methods revealed that the optimal values are lower than the estimated ones. The results of the uncertainty analysis of six hydraulic parameters showed that the uncertainty of the velocity is higher than the others. According to the results, the uncertainty is high, therefore, it is recommended MRC is determined with sufficient accuracy to reduce the financial costs and human losses caused by floods.